Professional Documents
Culture Documents
This book brings an important contribution for our understanding of how climate
change and variability impact society and livelihoods and in particular on how
the poor are affected. It should be read by everyone who has a responsibility in
development policies and in capacity building in LAC.
—Antonio Magalhães, Former Secretary of Planning, Brazil
Democratization and institutional reform and strengthening was, during the last
three decades, the Latin American way to respond to the challenges of coping
with severe macroeconomic imbalances, recovering growth, and reducing
poverty. This important book shows that this process has to be deepened to deal
with the social consequences of climate change.
—Andre Urani, President of Instituto Natura, Brazil
It is right to focus on the impact of climatic change on poor people. We all need to
be cognizant of the fact that those who have contributed the least to the deple-
tion of natural resources and to the changes in climate patterns are those who
will suffer the most from unsustainable human ecosystems.
—Lauritz B. Holm-Nielsen, Rektor at University of Aarhus, Denmark
Reducing Poverty, Protecting
Livelihoods, and Building
Assets in a Changing Climate
Social Implications of Climate Change in
Latin America and the Caribbean
Dorte Verner, Editor
© 2010 The International Bank for Reconstruction and Development / The World Bank
1818 H Street NW
Washington DC 20433
Telephone: 202-473-1000
Internet: www.worldbank.org
E-mail: feedback@worldbank.org
All rights reserved
1 2 3 4 13 12 11 10
This volume is a product of the staff of the International Bank for Reconstruction and
Development / The World Bank. The findings, interpretations, and conclusions
expressed in this volume do not necessarily reflect the views of the Executive
Directors of The World Bank or the governments they represent. The World Bank does
not guarantee the accuracy of the data included in this work.
The maps in this book were produced by the Map Design Unit of The World Bank.
The boundaries, colors, denominations, and any other information shown on these
maps do not imply, on the part of The World Bank Group, any judgement on the legal
status of any territory, or any endorsement or acceptance of such boundaries.
Rights and Permissions
The material in this publication is copyrighted. Copying and/or transmitting portions
or all of this work without permission may be a violation of applicable law. The
International Bank for Reconstruction and Development / The World Bank encourages
dissemination of its work and will normally grant permission to reproduce portions of
the work promptly.
For permission to photocopy or reprint any part of this work, please send a request
with complete information to the Copyright Clearance Center Inc., 222 Rosewood
Drive, Danvers, MA 01923, USA; telephone: 978-750-8400; fax: 978-750-4470;
Internet: www.copyright.com.
All other queries on rights and licenses, including subsidiary rights, should be addressed
to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC
20433, USA; fax: 202-522-2422; e-mail: pubrights@worldbank.org.
ISBN: 978-0-8213-8238-7
eISBN: 978-0-8213-8378-0
DOI: 10.1596/978-0-8213-8238-7
Cover photos: Dorte Verner
Cover design: Naylor Design, Inc.
Library of Congress Cataloging-in-Publication Data has been requested.
Contents
Foreword xv
Acknowledgments xvii
Abbreviations xix
Chapter 1 Introduction 1
Dorte Verner
v
vi Contents
Mexico 374
Central America 376
The Caribbean 381
South America 388
References 400
Index 415
Boxes
1.1 The Melting Glaciers in Bolivia: A Threat to People’s
Water Supply 2
1.2 The Sustainable Livelihoods Framework (SLF) 13
2.1 Water Scarcity and Climate Change Impacts
in The Bahamas 26
2.2 Water Scarcity, Climate Change, and Small Island
Dynamics in St. Vincent and Bequia 40
3.1 Flooding in Bolivia Caused by El Niño
Phenomenon 65
3.2 Natural Asset Erosion in Northeast Brazil 79
3.3 Social Capital Is Important in Natural
Disaster Management 81
x Contents
Figures
1.1 Climate Change and Its Social Implications 6
1.2 Changes in Water Runoff for the LAC Region
by 2050 Due to Climate Change 10
1.3 Key Climate Change Hot Spots for Latin America 11
3.1 Climate Impacts of El Niño Phenomenon in
Latin America and the Caribbean, 1998 67
4.1 Water and Wind Erosion and Desertification as
Sources of Land Degradation in Latin America and
the Caribbean 99
5.1 Impacts of Climate Change and Variability on the
Tourist Sector in Latin America and the Caribbean 137
5.2 Effects of Climate Change on the Economy, the
Tourist Industry, and Livelihoods 144
6.1 Main Effects of Climate Change on Health 170
Contents xi
Tables
1.1 Definition of Livelihood Assets 14
2.1 Supply and Demand for Renewable Freshwater
Resources in the LAC Region 23
2.2 Water Scarcity Levels 28
2.3 Climate Change Impact on Selected Major Cities in
Latin America and the Caribbean with Severe Existing
Water Problems 30
2.4 Rural Poverty in Countries with Arid and
Semiarid Regions 32
3.1 Frequency of Weather-Related Disasters (Floods,
Droughts, and Storms) in the LAC Region, 1970–2008 70
3.2 Impact of Weather-Related Disasters (Floods, Droughts,
Storms, and Extreme Temperatures) on People in the LAC
Region, by Subregion, 1970–2008 71
3.3 Weather-Related Disaster Exposure Indicators
in Latin America and the Caribbean, by Subregion,
1970–2008 71
3.4 People Killed or Affected by Weather-Related Disasters
in the LAC Region, 1970–2008 72
3.5 People Injured by Weather-Related Disasters in the LAC
Region, 1970–2008 73
4.1 Principal Production Systems among Agrarian
Rural Poor 96
5.1 Status of Coral Reefs in Latin America and
the Caribbean 126
5.2 Overview of the Fishing Industry in Select LAC
Countries 132
5.3 Sensitivity to Climate Change Based on Poverty Level
and Number of Fishermen 133
Contents xiii
Foreword
This book provides a much needed look at the impact of climate change
on the poor. It convincingly demonstrates that issues of poverty and
livelihoods must be integrated into climate change policies to help
achieve sustainable development gains.
The high incidence of natural disasters, growing urbanization, and
increased water scarcity—combined with the acute impact of these
phenomena on the poor and vulnerable—complicates the already enor-
mous challenge of reducing poverty and inequality in Latin America
and the Caribbean. This publication lays bare the social implications of
climate change and equips the reader with a framework for understand-
ing how climate change and climate variability affect livelihoods,
poverty, income, health, and migration.
Based on a study carried out by a multidisciplinary team in the
Sustainable Development department, this book’s timely analysis com-
plements the regional flagship report Low Carbon, High Growth. It
describes much-needed policy options on climate change adaptation.
Without appropriate adaptive responses, the impacts of climate
change on the vulnerable will be severe in the region. Key findings show
that changes in water flows brought about by climate change could
place 70 percent of the Andean population at risk of fresh water
xvi Contents
Pamela Cox
Vice President
Latin America and Caribbean Region, World Bank
Acknowledgments
The report and team were developed and managed by Dorte Verner
(Team Leader). The team members included Lykke Andersen, Inger
Brisson, Jens Hesselbjerg Christensen, John Bjerg Geary, Jakob Kronik,
Lotte Lund, Sara Trab Nielsen, Jørgen Eivind Olesen, Benjamin Orlove,
Claus Pörtner, Tine Rossing, and Olivier Rubin, and Sanne Agnete
Tikjøb.
Special thanks to the team of advisers: Jocelyne Albert, Shelton
Davis, Estanislao Gacitua-Mario, Andrew Norton, Walter Vergara, and
Alonso Zarzar.
The team is grateful to peer reviewers Kirk Hamilton and Andrea
Liverani for comments and suggestions. It also gratefully acknowledges
helpful comments and suggestions from Anjali Acharya, Arun Agrawal,
Maximilian Shen Ashwill, James Azueta, Herman Belmar, McDonald
Benjamin, Winston Bennett, Carter Brandon, Rita Cestti, Alejandro Deeb,
Pablo Fajnzylber, Dennis Garbutt, Maninder Gill, Jose Gutierrez, Gillette
Hall, Marea Hatziolos, Rasmus Heltberg, Willem Janssen, Ottis Joslyn,
Michel Kerf, Pilar Larreamendy, Celia Mahung, Alexandre Marc, Melanie
McField, Robin Mearns, Augusta Molnar, Edmundo Murrugarra, John
Nash, Frode Neergaard, Nicolas Perrin, Howie Prince, Gustavo Santiel,
Emmanuel Skoufias, and Senator Eddie Webster, of Stann Creek District,
xvii
xviii Acknowledgments
Belize. The team is grateful to the Latin America and Caribbean Region
management team and especially to Pamela Cox, Laura Tuck, McDonal
Benjamin, and Maninder Gill.
The team would also like to thank Ramon Anria and Jorge Hunt for
managing all the paperwork very effectively and Rachel Weaving for
editing the study. Moreover, the team would like to thank Pat Katayama,
Rick Ludwick, and Denise Bergeron, of the World Bank Office of the
Publisher, and Jeff Lecksell, of the Bank’s General Services Division,
Cartographic Services. Finally, the World Bank is grateful for financial
support from DANIDA, Denmark.
Abbreviations
Introduction
Dorte Verner
1
2 Verner
South America and western Central America are seeing decreases in pre-
cipitation and an increase in droughts. The Andean intertropical glaciers
are shrinking and are expected to disappear altogether within the next
20 to 40 years. Although this process will lead to increased water flow in
the short term and indeed pose a risk of glacier lake outbursts and
flooding, in the longer term the glaciers’ disappearance will severely cur-
tail the supply of water for drinking, irrigation, and power generation
(box 1.1). For Latin America and the Caribbean (LAC) as a whole, esti-
mates of the cost of damage due to climate change and climatic variabil-
ity vary from 1.3 percent to 7 percent of gross domestic product (GDP)
Box 1.1
the regulatory function they currently perform of evening out the flow of water
across the year will also cease. Indeed, unless measures are taken to capture runoff,
additional water may be lost, as the glaciers will no longer be there to absorb pre-
cipitation as it falls and release it slowly. Water will run off the mountains into the
ocean at greater speed than it does now.
Water quality is also an issue. El Alto has no sewerage treatment system, so a
smaller water supply means a higher concentration of waste. The resulting threat
to health will intensify. Diarrhea is already a serious problem as a result of contam-
inated water and food (for example, meat and vegetables grown outside La Paz
that have been irrigated by contaminated river water).
The melting of the glaciers also threatens agriculture and the energy supply
from hydroelectric plants. The implications are dire, not just for La Paz, but also for
Quito and Bogota, two other large cities that depend on glacial water. More than
11 million people now live in these three Andean cities, and the population of
El Alto alone is growing by 3 percent to 5 percent annually.
Source: Information gathered by the author during fieldwork in Bolivia in March 2008.
on the social dimensions of climate change in the region and the pol-
icy options for addressing them.
Much of this book relies on new empirical work. The authors undertook
quantitative analyses of relationships in LAC countries between climate
change and (a) health indicators (life expectancy and child mortality),
(b) income or consumption and inequality, and (c) migration, using
municipal-level data. Research included extensive field surveys in indige-
nous communities across the region to investigate how individuals and
communities perceive and respond to climate change and variability
(complete analyses can be found in Kronik and Verner 2010).
The book focuses on climate change’s social impact on the most vul-
nerable population groups in the LAC region and on necessary ele-
ments of adaptation measures that can reduce the negative effects that
they will otherwise face. The book does not address mitigation or any
negative social impacts potentially associated with it, despite their
obvious importance.5 Neither does the book address issues of energy,
forests, or infrastructure, since those areas are the subjects of other
ongoing work.6
The book’s focus is on the most vulnerable people, who tend to be
the poorest.7 The poor are more vulnerable than the nonpoor because
they depend more heavily on natural resources for their livelihoods and
well-being and have limited capacity to cope with unpredictable and
extreme weather. Traditional knowledge about their environment—a
key asset of indigenous peoples and poor farmers, in particular—may no
longer be reliable in the face of climatic changes. In addition, the poor
tend to live in areas at greater risk from extreme weather events (for
example, on flood plains, unstable slopes, or unproductive land) and to
lack the assets necessary to adapt to rapidly changing circumstances.
Some poor and vulnerable people share a strong sense of injustice,
believing that state policies have not benefited them. A large proportion
of poor and vulnerable people identify insecurity, income inequality,
and lack of political representation as critical factors reinforcing the
vicious cycle of deprivation and destitution that they find themselves
in. Others, including many indigenous people, blame the effects of cli-
mate change on their own failings.8
The progression of cause and effect underlying this book is sketched
in figure 1.1, showing how greenhouse gas emissions resulting from
human activity are linked to their environmental impacts (step 1). This
environmental degradation affects the availability and quality of water
for human consumption (including domestic, agricultural, and industrial
6 Verner
use and power generation), as well as terrestrial and marine flora and
fauna ecosystems. The environmental impacts have social implications, as
presented in step 2, affecting people’s livelihoods, food security, and
health. Excessive stress on those determinants of human well-being will
increase poverty and income inequality. It may also cause migration to
swell and has the potential to heighten the risk of conflict.
Although many of the effects of climate change and variability are
already unavoidable, much scope remains for human agency and ingenu-
ity in crafting strategies to mitigate them by addressing the causes of cli-
mate change itself and for adaptation to address the consequences.9 An
optimal national strategy would employ both mitigation and adaptation
efforts and should embody good governance, providing for the exercise of
voice, representation, and social accountability.
It is critical that policy makers in Latin America and the Caribbean
address the social issues related to climate change. Even if global mitiga-
tion efforts improve, the climate trends that are already under way have
considerable momentum and will dramatically affect economic, human,
and social development for years to come. Thus, as the Stern Review on the
Introduction 7
the changes and to identify ways to help them increase their resilience
and adaptive capacity.11
Extreme events. Most areas of Latin America and the Caribbean have
experienced several instances of severe weather in recent decades, with
torrential rain and hurricanes causing thousands of deaths and damaging
properties, infrastructure, and natural resources. These events are widely
interpreted as reflecting climate change and increased variability, but no
formal scientific detection of such a relationship at the regional level has
been made.13 Neither is there evidence that the extreme events recently
seen are less severe, or more so, than those that may be experienced in the
future. This said, the available models clearly suggest that changes will
Figure 1.2 Changes in Water Runoff for the LAC Region by 2050 Due
to Climate Change
More than 20
20 to 0
0 to –20
–20 and more
IBRD 37780
MAY 2010
Source: Author’s extraction of the LAC region from global map (Arnell 2004).
Introduction 11
take place over this century that will generally be in the direction of more
of the extremes, that is, more intensive precipitation, longer dry spells and
warm spells, heat waves with higher temperatures than generally experi-
enced up to now, and more numerous severe hurricanes. According to
the Intergovernmental Panel on Climate Change (IPCC 2007a), further
increases are expected in floods and droughts and in the intensity of trop-
ical cyclones. Regional hot spots are shown in figure 1.3.
The countries in the Caribbean and the Gulf of Mexico, which are
often assailed by intense hurricanes, can expect these storms to become
even fiercer as a result of climate change. Other important issues are the
destruction of coral reefs and a growing threat to southeast Pacific fish
stocks due to increasing sea surface temperatures. A rise in sea level
would likely bring flooding to low-lying regions such as the coasts of El
Salvador and Guyana and could exacerbate social and political tensions
in the region.
Figure 1.3 Key Climate Change Hot Spots for Latin America
IBRD 37781
MAY 2010
Analytical Framework
To permit a systematic analysis of the effects of climate change on the
poor and vulnerable, a slightly adapted version of the United Kingdom
Department for International Development’s sustainable livelihoods
framework (SLF), described in box 1.2, is used (DFID 2001). This frame-
work makes it possible to see how different aspects of climate change and
climatic variability affect people’s assets, their livelihood strategies and
livelihood outcomes, and hence their well-being. Thus it also sheds light
on possible entry points for efforts to raise income, increase well-being,
reduce vulnerability, improve food security, or achieve more sustainable
use of resources.
Introduction 13
Box 1.2
Livelihood Assets
H
Vulnerability N P Influence Transforming Livelihood Livelihood
Structures
Context Strategies Outcomes
F C & Access and Processes
S
access to assets. For this book, the authors have added cultural capital
(Bourdieu 1973, 1986; Bourdieu, de Saint Martin, and Clough 1996)
to the five livelihood assets considered in DFID’s original SLF. Field
research revealed that, particularly for indigenous people, the cultural
dimension of livelihood strategies and social institutions is important for
understanding the impacts of climate change and climatic variability.
The SLF is a tool for assessing the vulnerability of the poor and their
capacity to cope with shocks and adapt to changing trends—all highly
important from a climate change perspective. Vulnerability, and the
added vulnerability resulting from climate change and climatic variability,
can be assessed for different population groups, as can people’s ability to
adapt to climate change within their specific environmental context.
Notes
1. “Climate change” refers to long-term change, that is, slow changes in, for
example, mean annual temperature and precipitation levels, leading to sea
level rise, melting of glaciers, and so on. “Change in climatic variability” refers
to increased unpredictability with respect to seasons, rainfall, hurricanes, and
the like, and is largely related to the El Niño-Southern Oscillation (ENSO).
2. WRI 2005. The figure falls to 6 percent of global emissions if emissions from
energy use only are taken into account. When land use changes are included, the
proportion rises to 12 percent of the world’s green house gas (GHG) emissions,
mainly as a result of the large-scale deforestation taking place in the region. Yet,
despite the region’s small contribution to global warming, its people still find
their well-being, homes, and livelihoods threatened by climate change and cli-
matic variability. The inverse relationship between responsibility for global
warming and vulnerability to its effects is often ignored (UNDP 2007).
3. These estimates are higher than earlier ones that only included agricultural
losses for the region, ranging from US$ 35 billion per year, equivalent to 0.23
percent of gross domestic product (GDP) (Mendelsohn and Williams 2004),
to US$ 120 billion per year, equivalent to 0.56 percent of GDP (Tol 2002),
by 2100. A more recent study projects total losses for the region of about US$
91 billion, equivalent to around 1 percent of GDP, by 2050, if warming
reaches 1.79°C relative to 1900 (Medvedev and Van Der Mensbrugghe
2008). This estimate is also lower than the range estimated by Nagy and oth-
ers (2006), but it excludes noneconomic sectors. It also does not take account
of the effects of possibly increased frequency and severity of natural disasters
as a result of climate change or of potentially catastrophic climate change
effects, such as collapsing ice sheets or melting permafrost.
4. The Intergovernmental Panel on Climate Change’s landmark Fourth
Assessment Report, released in 2007, provides scientific evidence that emissions
from human activity, particularly burning fossil fuels for energy, are causing
changes to the Earth’s climate (IPCC 2007c). The Stern Review examined the
economics of climate change (Stern 2007). Other authoritative sources
include the Human Development Report 2007/2008 (UNDP 2007); Mathur,
Burton, and van Aalst 2004; and De la Torre, Fajnzylber, and Nash 2009.
5. Reducing emissions from deforestation and degradation (REDD) and other
mitigation measures, energy policies, and infrastructure can all have positive or
negative social impacts, depending on their design. While much effort is
invested into finding viable ways to mitigate greenhouse gases, only recently
have the social effects of mitigation measures surfaced as an area requiring
attention. A few examples illustrate these complex relations. Switching to
renewable energy sources, such as hydropower or biofuels, can have negative
social implications unless attention is paid to how the switch takes place. Thus,
building dams for hydropower often displaces large numbers of people, and
Introduction 17
reducing the social costs requires close consultation and cooperation with
and among the affected people to ensure that the relocation takes into
account their wishes and needs, including for livelihood. Similarly, large-scale
cultivation of biofuels may displace subsistence farmers from their land,
forcing them either to move to less-productive land or to clear forest for
new land. At the same time, a switch to growing biofuels may displace food
crops, resulting in food price hikes that threaten the food security of the
poor. “Hard” adaptation measures, such as sea defenses and other infrastruc-
ture, can also affect vulnerable segments of the population in positive or
negative ways, depending on their configuration.
6. Given the scale of these areas, the World Bank is undertaking work that cov-
ers the forest sector and REDD exclusively and similarly covers the energy
and infrastructure sectors extensively. It would seem a natural follow-up to
this volume to initiate a similar study of the social impacts associated with
mitigation. A number of countries within LAC may introduce REDD pro-
grams in an effort to reduce greenhouse gas emissions and receive credits for
it. However, any effort to nationalize forests must be considerate of indige-
nous peoples, who may fear the impact on their livelihoods. Forming REDD
partnerships recognizing these peoples’ autonomy and authority can be a
strategy for effective and sustainable forest management.
7. The effects of climate change have very unequal distributional implica-
tions and as such are socially differentiated. Hence some have called cli-
mate change the “greatest social injustice of our time” (as stated by Mary
Robinson, International Institute for Environment and Development, on
December 11, 2006).
8. Interviews of indigenous people for this book; see also Kronik and Verner
2010.
9. The essence of mitigation policies should be to price carbon and carbon
equivalent gases so as to reflect their true costs—including social costs. This
book does not address these policies in any detail.
10. Climate change assumptions are detailed in appendixes A and B, which sum-
marize existing knowledge about recent and most likely future climate change
and variability in the LAC region, building on the recent Intergovernmental
Panel on Climate Change (IPCC) Fourth Assessment Reports (2007a; 2007b;
2007c) and recent literature.
11. Appendix C provides further details of the analytical framework and method-
ology underlying the book.
12. There will be local exceptions to these broad tendencies, but existing models
do not give robust results for every part of this large region of interest. In par-
ticular, there are still many unresolved issues related to changes in the
Amazonas, as important aspects of the interaction between vegetation and cli-
mate are still not well understood.
18 Verner
References
Agrawal, Arun. 2008. “The Role of Local Institutions in Adaptation to
Climate Change.” Paper presented at World Bank Workshop on Social
Dimensions of Climate Change, Social Development Department, World
Bank, Washington, DC.
Arnell, N. W. 2004. “Climate Change and Global Water Resources: SRES Scenarios
and Socio-economic Scenarios.” Global Environmental Change 14: 31–52.
Bourdieu, Pierre. 1973. “Cultural Reproduction and Social Reproduction.” In
Knowledge, Education, and Cultural Change, ed. Richard Brown. London:
Willmer Brothers Ltd.
———. 1983, 1986. “The Forms of Capital.” Translated into English by Richard
Nice. In Handbook of Theory and Research for the Sociology of Education, ed.
J. G. Richardson, 241–58. New York: Greenwood Press.
Bourdieu, Pierre, Monique de Saint Martin, and Laurette C. Clough. 1996. The
State Nobility. Stanford, CA: Stanford University Press.
Buhaug, H., N. Gleditsch, and O. Theisen. 2008. “Implications of Climate Change
for Armed Conflict.” Paper presented at World Bank Workshop on Social
Dimensions of Climate Change, Social Development Department, World
Bank, Washington, DC.
De la Torre, A., P. Fajnzylber, and J. Nash. 2009. Low Carbon High Growth: Latin
American Responses to Climate Change. An Overview. World Bank Latin
American and Caribbean Studies. Washington, DC: World Bank.
DFID (Department for International Development, UK). 2001. Sustainable
Livelihoods Guidance Sheet. http://www.nssd.net/pdf/sectiont.pdf.
IPCC (Intergovernmental Panel on Climate Change). 2007a. Climate Change
2007: The Physical Science Basis. Contribution of Working Group I to the
Fourth Assessment Report of the IPCC. Geneva: IPCC.
———. 2007b. Climate Change 2007: Impacts, Adaptation, and Vulnerability.
Contribution of Working Group II to the Fourth Assessment Report of the
IPCC. Geneva: IPCC.
Introduction 19
Few resources have more critical bearing on human livelihoods and well-
being than water. Water is essential for all socioeconomic development
and for maintaining healthy ecosystems. Throughout history, the success
or failure of societies in harnessing the productive potential of water,
while limiting its destructive potential, has determined human progress.
This chapter examines how climate change and variability will affect
water scarcity and thereby human livelihoods in Latin America and the
Caribbean (LAC)—a region in which about 60 percent of the population
is concentrated in 20 percent of the land area, largely in arid and semiarid
coastal and mountainous zones that hold only 5 percent of the region’s
water resources (WMO/IADB 1996).
Water scarcity is relative. In general, it is defined as “the point at which
the aggregate impact of all users impinges on the supply or quality of
water under prevailing institutional arrangements to the extent that the
demand by all sectors, including the environment, cannot be satisfied fully”
(UN Water 2007). As might be expected, water scarcity is often rooted in
water shortage; it tends to be most acute in arid or semiarid regions that
are affected by droughts and climatic variability, combined with rapid pop-
ulation growth and economic development. However, water scarcity can
also be a social construct, reflected in, for example, imbalances between
21
22 Rossing
Access to
improved Access to
Freshwater supply Freshwater demand water improved
Population (Actual renewable water resources)1 (Total annual water withdrawals)1 source2 sanitation2
(summed by sector)
Surface water: total
Groundwater: total
renewable (actual)
renewable (actual)
renewable (actual)
Urban population
Urban population
Rural population
Rural population
Total population
Withdrawals by sector
with access
with access
with access
with access
Agriculture
in volume
Per capita
Domestic
Industry
LAC
Sub-Region / (1,000 (1,000 (1,000 (1,000 (1,000 (m3/ (1,000 (1,000
Country inhab) m3/yr) m3/yr) m3/yr) m3/yr) inhab/yr) m3/yr) m3/yr) (%) (%) (%) (%) (%) (%) (%) (%)
Central
America 143,775 2,650 1,036 335 1,165 25,802 87.3 378 60.2 18 21.9 3.7 80 96 60 81
Belize 276 39 19 68,722 0.2 593 20 73.3 6.7 0.8 82 100 25 71
Costa Rica 4,327 150 75 37 112 25,976 2.7 654 53.4 17.2 29.5 2.4 96 99 95 96
El Salvador 6,668 36 25 6 25 3,667 1.3 196 59.4 15.6 25.0 5.1 68 94 80 90
Guatemala 12,710 217 103 34 111 8,832 2.0 172 80.1 13.4 6.5 1.8 94 99 79 90
Honduras 6,834 221 87 39 96 13,314 0.9 128 80.2 11.6 8.1 0.9 74 95 55 78
Mexico 104,266 1,472 409 139 457 4,272 78.2 760 77.1 5.48 17.4 17.0 85 98 48 91
Nicaragua 5,463 311 193 59 197 35,847 1.3 252 83.1 2.31 14.6 0.7 63 90 34 57
Panama 3,232 203 145 21 148 45,786 0.8 268 28 4.88 67.1 0.6 81 96 63 78
23
(continued)
Table 2.1 Supply and Demand for Renewable Freshwater Resources in the LAC Region (2005 or Most Recent Data) (continued)
24
Access to
improved Access to
Freshwater supply Freshwater demand water improved
Population (Actual renewable water resources)1 (Total annual water withdrawals)1 source2 sanitation2
(summed by sector)
Surface water: total
Groundwater: total
renewable (actual)
renewable (actual)
renewable (actual)
Urban population
Urban population
Rural population
Rural population
Total population
Withdrawals by sector
with access
with access
with access
with access
Agriculture
in volume
Per capita
Domestic
Industry
LAC
Sub-Region / (1,000 (1,000 (1,000 (1,000 (1,000 (m3/ (1,000 (1,000
Country inhab) 3 3
m /yr) m /yr) m /yr) m /yr) inhab/yr) m3/yr) m3/yr)
3 3 (%) (%) (%) (%) (%) (%) (%) (%)
South
America 373,679 28,375 17,084 3693 17,132 89,185 164.7 789 74.5 8.2 17.3 1.3 75 97 60 87
Argentina 38,747 1,642 814 128 814 21,008 29.2 775 73.7 9.46 16.8 3.6 80 98 83 92
Bolivia 9,182 1,259 596 130 623 67,799 1.4 166 80.6 6.88 12.5 0.2 69 96 22 54
Brazil 186,831 15,236 8,233 1,874 8,233 44,167 59.3 331 61.8 18 20.3 0.7 58 97 37 84
Chile 16,295 1,152 922 140 922 56,582 12.6 796 63.5 25.2 11.3 1.4 72 98 74 97
Colombia 44,946 2,975 2,132 510 2,132 46,754 10.7 246 45.9 3.73 50.3 0.5 77 99 58 85
Ecuador 13,061 592 424 134 424 32,083 17.0 1,340 82.2 5.3 12.5 4.0 91 98 72 91
Guyana 739 513 241 103 241 320,905 1.6 2,195 97.6 0.61 1.8 0.7 91 98 80 85
Paraguay 5,904 460 336 41 336 54,563 0.5 85 71.4 8.16 20.4 0.1 52 94 42 89
Peru 27,274 2,234 1,913 303 1,913 68,400 20.1 752 81.6 10.1 8.4 1.1 63 92 36 85
Suriname 452 381 122 80 122 271,715 0.7 1,519 92.5 2.99 4.5 0.5 79 97 60 89
Uruguay 3,326 223 139 23 139 40,139 3.2 929 96.2 1.27 2.5 2.3 100 100 99 100
Venezuela 26,726 1,710 1,211 227 1,233 46,102 8.4 330 47.4 7.05 45.5 0.7 72 93 51 91
Caribbean 40,525 328 70 24 94 1,843 13.4 331 51 17 32.0 28.4 86 94 82 85
Bahamas 323 18 ~ ~ 0 62 ~ ~ ~ ~ ~ ~ 100 98 100 100
Barbados 292 1 0 0 0 296 0.1 336 22.2 44.4 33.3 113.0 86 100 100 99
Cuba 11,260 148 32 6 38 3,383 8.2 733 68.8 12.2 19.0 21.5 78 95 95 99
Dominican
Republic 9,470 69 21 12 21 2,360 3.4 398 66.1 1.77 32.2 16.1 91 97 74 81
Haiti 9,296 40 12 2 14 1,645 1.0 121 93.9 1.01 5.1 7.1 51 70 12 29
Jamaica 2,682 23 6 4 9 3,547 0.4 157 48.8 17.1 34.1 4.4 88 97 84 82
Puerto Rico 3,947 18 ~ ~ 7 1,795 ~ ~ ~ ~ ~ ~ ~ ~ ~ ~
S. Kitts and
Nevis 49 1 0 0 0 558 ~ ~ ~ ~ ~ ~ 99 99 96 96
Trinidad and
Tobago 1,324 11 ~ ~ 4 2,943 0.31 240 6.45 25.8 67.7 8.07 93 97 92 92
Total LAC 557,97 31,353 18,190 4,052 18,391 38,943 265.4 499 62 14.4 23.6 11 86 94 82 85
Sources: UN Population Division2006; World Bank 2005; Quijandria, Monares, and de Peña Montenegro 2001; FAO Aquastat Water Data 2008; and World Bank LAC Water and Sanitation, Data and
Statistics Website. The most recent data are used and stem from 2005–07, but others are long-term averages from multiple sources and years.
Notes: 1) FAO (2008); 2) World Bank LAC Water – Data and Statistics website. The most recent data are used and most stem from 2005–2007, but other are long-term averages from multiple
sources and years.
25
26 Rossing
Box 2.1
Water losses are great. For New Providence, water loss is estimated at 53 percent,
roughly equivalent to the amount barged in from Andros. In addition, as a result of
overextraction of the limited freshwater reserves, saltwater intrusion is already oc-
curring on New Providence, which has the country’s greatest demand for water.
The nature of the geology and the lack of proper sewage collection and treatment
systems also contribute to the contamination of groundwater.
(continued)
Water Scarcity, Climate Change, and the Poor 27
The country’s freshwater supplies are highly vulnerable to the effects of cli-
mate change. Being very shallow, the aquifers are at great risk of becoming inun-
dated with saline water if even a small rise in sea level occurs. Less precipitation
over the years in some islands as a result of climate change is reducing freshwater
availability. Natural disasters and severe weather, such as hurricanes, are probably
the greatest threat to the health of the freshwater reserves. Once polluted,
groundwater is very expensive to clean up. Protecting it from contamination is
preferable and more cost-effective than remediation. Little, if anything, can be
done to protect the groundwater from natural disasters, however. As a result of
these and other factors, reverse osmosis (RO) is a key to future water supply for
New Providence and many other islands in The Bahamas, particularly the central
and southernmost islands. At present, Grand Bahama, Abaco, and Andros have
enough fresh groundwater reserves to meet their demands.
Since water is a limiting factor for economic and social development, The
Bahamas should treat groundwater as a strategic national resource. Regulating
and protecting its water resources is essential, and the use of integrated
groundwater management for this purpose is recommended. Current laws
and regulations, particularly those governing land use, planning, and water
management, are unclear and inadequate. Ignoring the overexploitation of water
resources will have severe repercussions, including health risks from waterborne
diseases and much greater water costs. Water will become much more costly if it
has to be treated because of groundwater contamination, if RO needs to be used,
or if more water has to be barged to meet demand. Proper land use planning
and regulations, currently lacking, must play an important role in protecting this
resource. The Ministry of Health and the Environment has proposed creating a
new department of environmental planning and protection to regulate ground-
water extraction and control pollution.
Sources: U.S. Army Corps of Engineers 2004; Bahamas Water and Sewerage Corporation 2008,
http://www.wsc.com.bs/waterandsewerage.asp.
Table 2.3 Climate Change Impact on Selected Major Cities in Latin America and the Caribbean with Severe Existing Water Problems
Share of
urban HH Share of
Main threats Population, connected urban HH Water Primary water source
from climate millions to water connected to demand Surface Surface + Existing water quality
change City Country (year) (%) sewage (%) (m3/sec) water Groundwater groundwater problems
Glacier melt Bogota Colombia 5.1 ~ ~ 17 X Bogota River highly
(1995) polluted by draining
urban wastewaters.
El Alto Bolivia 0.7 33 20 ~ X La Paz River highly
(1998) polluted from draining
urban wastewaters.
La Paz Bolivia 1.2 55 58 ~ X La Paz River highly
(1998) polluted from draining
urban wastewaters.
Glacier melt, Lima Peru 7.1 70 69 25 X Overexploitation of
coastal (1995) aquifers leading to
inundation, saline intrusion; 2,000
sea level rise hectares of vegetable
crops irrigated with
urban wastewaters.
Coastal Buenos Aires Argentina 12.6 ~ ~ 85 X Highly polluted surface
inundation; (1995) water from urban
sea level rise wastewaters draining
into the Riachuelo.
Montevideo Uruguay 1,6 ~ ~ 5 X Pollution; limited
(1995) resources.
Drought Guatemala Guatemala 1.1 ~ ~ 5 X Location on a divide;
City (1995) lack of water.
Mexico City Mexico 22.8 ~ ~ 50 X Severe groundwater
(1995) overuse and
contamination; about
90,000 hectares of
agricultural land in the
Tula Valley irrigated
with wastewater from
the city.
Sao Paulo Brazil 16.8 ~ ~ 50–55 X Contaminated waters
(1995) from the Tiete River
used to irrigate
vegetable gardens
downstream from
the urban core.
Santiago Chile 4.8 98 92 20 X Mapocho River highly
(1995) polluted from draining
urban wastewaters;
contaminated
wastewaters used
for irrigation.
Sources: Author’s preparation based on data from Alfaro and Marin 1994; Anton 1993; Fernandez and Graham 1999; Ringler, Rosegrant, and Paisner 2000; and WRI 1998.
Note: HH = household; m3/sec = cubic meters per second; ~ = no data available.
31
32 Rossing
Table 2.4 Rural Poverty in Countries with Arid and Semiarid Regions
Total rural population Rural population below the
Country 2005 (1,000) national poverty line (%)
Brazil 29,462 41.0
Honduras 3,657 70.4
Mexico 24,702 27.9
Nicaragua 2,407 64.3
Peru 7,881 72.1
Venezuela 2,051 n.a.
Source: UN Population Division. 2006; World Bank 2008b.
Note: n.a. indicates not available.
Water Scarcity, Climate Change, and the Poor 33
third (33.6 percent) of the urban poor have access to a flush toilet, but
only 12.6 percent of the rural population do (National Research Council
2003).6 Notably, access is much more limited in smaller cities than in
larger ones. The greater ease of providing services to clustered rather than
scattered populations helps explain some of the service differences
between the urban and rural poor (Fay 2005).
Despite the better services in urban areas, the basic service needs of the
urban poor are seldom fully met, for several reasons. First, coverage
remains far from universal. Even in a relatively wealthy country such as
Argentina, 47 percent of the urban poor lacked adequate sanitation in
1998 (World Bank 2000). In Cali, Colombia, a fifth of the poorest did not
have access to a private toilet in 2000 (World Bank 2002). And half of the
favela dwellers in Rio de Janeiro, Brazil, lived without a sewerage connec-
tion in 2000 (World Bank 2001). Second, quality and reliability tend to
be inadequate, and services are often unaffordable. In Guayaquil,
Ecuador, about 40 percent of the city’s population has to make do with
only 3 percent of the piped water, and some 800,000 people living in
low-income or informal settlements depend on water vendors (UNDP
2006). Without access to safe water, the poor must often rely on bot-
tled water, paying 5 to 10 times more than households that have piped
water. Third, access to services also reflects disparities between indige-
nous and nonindigenous people. For example, in Bolivia the average rate
of access to piped water is 49 percent for indigenous people, compared
to 80 percent for nonindigenous people (UNDP 2006).
Because of crowding, enormous repercussions for public health—often
leading to decreased economic opportunities—may follow if even a frac-
tion of a neighborhood lacks access to sanitation. People who cannot
afford clean water use contaminated water, which, together with poor
hygiene practices and poor sanitation, greatly increases their exposure to
waterborne diseases (chapter 6). Often people are forced by poverty to
settle on any available land they can find. They live in great numbers
along water drains, river banks, and the like, using them for their sanitary
needs. They use the water that flows in them for washing and laundry, and
that both adds to the pollution and exposes them to it. In heavy rains
these water drains and rivers can turn into virtual death traps, as people’s
belongings and litter cause blockages that result in flash floods.
As water becomes scarcer, more of it will be withdrawn from low-
quality sources. More frequent heavy rainfall in some areas will increase the
pollution and sediment loads carried by streams and rivers through higher
runoff and infiltration. The associated risks include the malfunctioning of
Water Scarcity, Climate Change, and the Poor 35
water infrastructure during floods. Drainage systems are often built with
insufficient capacity to cope with the sudden, large volumes of water that
have been seen in recent years, and which may recur with greater fre-
quency, and that may allow contamination of drinking water and overload-
ing of water and wastewater treatment plants during extreme rainfall. The
problem of drinking water contamination is particularly severe in cities
where storm and sanitary sewers are not separate. These causal links help to
explain why infant and child mortality are higher among the urban than the
rural poor in a number of countries, despite better urban access to both
infrastructure and health services (Fay 2005).
Although urban water demands are still several times smaller than the
demands of irrigated agriculture, growing water scarcity in urban areas is
a consequence of the spatial concentration of urban demand. The popu-
lation of the LAC region continues to urbanize at some of the fastest rates
in the world. Urbanization in the region is expected to increase from its
current rate of 79 percent to 81 percent by 2015 (World Bank 2008a;
Ravallion, Chen, and Sangraula 2008). With this trend comes increased
pressure on urban water supply and sanitation systems. Unless poverty
rates decrease, population growth and continued urbanization imply an
additional 22 million urban poor in the LAC region by 2015, 9 million of
whom will be living in extreme poverty (Fay 2005). Notably, the urban-
ization is likely to be fastest in the least urbanized countries in the region
(the Central American countries, Bolivia, Ecuador, and Paraguay). Many
of them, such as Bolivia, are already struggling with water scarcity chal-
lenges (UN Water 2006).
Box 2.2
Bequians receive no government support for their water supply, and the
state-owned Central Water and Sewerage Authority (in St. Vincent) does not
even operate on the island. Consequently, even though the Bequian population
is poorer, they pay considerably more for their water than people in St. Vincent.
The cost of water shipped from St. Vincent during droughts and the dry seasons
is borne entirely by Bequians.
The prospect of climate change and variability is troublesome for Bequia since,
as the IPCC projects, summer rainfall in the Caribbean, especially in the vicinity of
the Greater Antilles, will likely decline. It is also unlikely that increased rainfall dur-
ing the winter months will compensate for the decline, inasmuch as storage facil-
ities in Bequia are modest at best. Add to this the likelihood of increased demand
due to population growth, and the risks of sea level rise and coastal flooding, and
the future looks extremely challenging for the people of Bequia.
Sources: Based largely on interviews and information gathered during a field trip for this study to
St. Vincent and Bequia. Other sources include World Bank 2006; National Environmental Advisory
Board/Ministry of Health and the Environment 2000; IPCC 2008; Oki and Kanae 2006.
Given that both water scarcity and the effects of climate change and
variability on water resources tend to be local, IWRM must be tailored to
specific local circumstances, mainly through the formulation of location-
specific policies and institutional arrangements. Solutions will also need to
consider the local dynamics of both water supply and demand, along with
the expected consequences of climate change and variability. For water
managers, it is crucial both to reduce the level of demand and to distrib-
ute supplies efficiently, while resolving conflicts of interest cooperatively
and allocating rights of access and use equitably. Moreover, because both
water scarcity and climate impacts are location specific, work is needed to
enhance the capacity of local authorities and communities to initiate and
implement adaptive measures and to develop lasting mechanisms that
enable them to work in partnership with one another. Such a locally
focused approach will help strengthen the sustainability of adaptation ini-
tiatives, while avoiding exclusion of the needs and priorities of marginal
groups, particularly the poor.
Latin America and the Caribbean are increasingly joining the global
trend toward IWRM and the delegation of water management responsi-
bility to local authorities and water user groups, thereby promoting the
principle of subsidiarity. Over the past 10 to 15 years, many river basin
management organizations have been created with the goal of facilitating
stakeholder involvement to achieve more integrated water management.
But while appropriate legal and institutional frameworks have been created
in a variety of places, the implementation of IWRM has met a number of
important obstacles. Initial conditions and different stakeholder capacities
and interests can have an important influence on the success of decentral-
ized water resources management.
The case studies below provide examples in which responsibility for
water management and adaptation to climate change and variability is
being devolved or decentralized from national governments and agencies to
regional or local government authorities. The cases are drawn from loca-
tions that have experienced—and will likely continue to experience—water
scarcity compounded by adverse climate-related impacts: northeast Brazil,
Mexico, and Bolivia.
the dry season, will force both national and municipal authorities to pay
considerable attention to including stakeholders in discussions of water-
related issues, to avoid serious unrest.
Some promising steps have already been taken.23 In 2002, the
Interinstitucional del Agua (CONAIG) was created (through Decree no.
26599) to provide a forum in which government, social, and economic
organizations could discuss legal, institutional, and technical aspects of
water resources management. The cornerstones of the new Agua para
Todos framework are
Policy Implications
The compounding impact of climate change on water resources increases
the urgency of using integrated planning approaches, as is already evident,
especially in arid regions. Both water managers and decision makers should
be encouraged to engage in greater dialogue with climate and develop-
ment specialists to better understand the climate-related challenges and
how to deal with them. A suggested first step in reducing vulnerability to
climate change would be a “no-regrets” approach of improving or encour-
aging adaptation to existing climate variability while also improving the
governance and management of water resources.
To function well, decentralized institutions need solid management
and technical expertise, as well as timely access to necessary funds.
Although it is crucial to establish the proper legal and institutional
frameworks, human and financial resources must also be considered in
the long-term planning and management of water resources at the
local level.
52 Rossing
advantage of improving data sets, they seek to address some of the limi-
tations of single-factor indicators. Future fieldwork would therefore ben-
efit from applying the new indexes to better assess the effects of climate
change for water resources and water systems.
A promising example is the Vulnerability of Water Systems index
developed by Peter Gleick.24 This index measures five aspects of vulner-
ability at the watershed or basin level: (a) storage volume relative to total
renewable water resources; (b) consumptive use relative to total renew-
able water resources; (c) proportion of hydroelectricity relative to total
electricity; (d) groundwater overdraft relative to total groundwater with-
drawals; and (e) variability of flow. The combination of these five aspects
makes it possible to measure both water problems and the effects of cli-
mate change at the local level. For example, the first measure is an indi-
cator of the ability of an area to withstand prolonged droughts or severe
flooding. With large storage compared to supply, short-term droughts are
less likely to cause major water shortages. In contrast, when this ratio is
small, changes in the intensity of floods and droughts may be more
strongly felt.
Another example is the Climate Vulnerability Index (CVI) being
developed by the Oxford University Centre for the Environment. It
provides a powerful tool to express systematically the vulnerability of
human communities in relation to water resources, using a holistic
approach that integrates the physical, social, economic, and environ-
mental factors. Notably, the CVI is suitable for examining vulnerabil-
ity to present levels of climate variability. It can also be used to
examine the impacts of climate change, combining climate scenarios
with expected changes in technological, social, economic, and environ-
mental conditions. Another promising feature is that the index’s out-
puts can be linked directly to impacts on people through application
of the sustainable livelihoods framework (SLF), which allows the
poverty implications to be examined. Tentative fieldwork from loca-
tions in South Africa, Sri Lanka, and Tanzania confirms that the poor
are the most vulnerable to climate impacts. Finally, the identification
of zones of vulnerability provides a systematic rationale for determin-
ing priority areas where proactive measures to protect populations
should be taken.25
Finally, to develop a set of best practices of use to the LAC region, it
would be interesting to focus future case studies on how IWRM can be
tailored to local circumstances and how scientific knowledge about cli-
mate change and IWRM principles is being applied on the ground.26
54 Rossing
Notes
1. For the past two decades, the Falkenmark Water Stress Index has been the
most influential and powerful water measure. Generally, 1,700 cubic meters
(m3) per person per year is considered the level above which water shortages
are rare and localized. Between 1,700 and 1,000 m3 per person per year is
considered regular water stress. Between 1,000 and 500 m3 per person per
year is considered chronic water scarcity, which begins to hamper health, eco-
nomic development, and well-being, and below 500 m3 per person per year
is considered absolute/severe water scarcity, where water availability is a pri-
mary constraint on life. For more elaboration on the Falkenmark Water Stress
Index, see Gleick and others 2002.
2. Notably, a figure for The Bahamas is not available.
3. The arid northwest and central regions contain 77 percent of Mexico’s popu-
lation and generate 85 percent of GDP (Wikipedia.org, “Mexico”).
4. See also World Bank Water and Sanitation Web site for info on the LAC region,
http://web.worldbank.org/WBSITE/EXTERNAL/TOPICS/EXTWAT/
0,,contentMDK:21706928~menuPK:4602430~pagePK:148956~piPK:216618
~theSitePK:4602123,00.html. Surface water and, to a lesser extent, groundwa-
ter in these cities has become contaminated with varying degrees of detergents
(soaps and solvents), pesticides, petroleum and its derivatives, toxic metals (such
as lead and mercury), fertilizers and other plant nutrients, oxygen-depleting
compounds (for example, waste from canneries, meat-processing plants, slaugh-
terhouses, and paper and pulp-processing operations), and disease-causing
agents responsible for hepatitis and infections of the intestinal tract such as
typhoid fever, cholera, and dysentery (Anton 1993).
5. World Bank Water and Sanitation Web site, http://web.worldbank.org/
WBSITE/EXTERNAL/TOPICS/EXTWAT/0,,contentMDK:22005375~page
PK:148956~piPK:216618~theSitePK:4602123,00.html.
6. These figures are obtained from a probit analysis. Poverty is identified as the
lowest quartile of a composite assets-and-durables index, as there are no mon-
etary values in the demographic and health surveys on which the analysis is
based. Data are from various years.
7. Peru’s 18 glaciers have lost 22 percent of their surface over the past 27 to
35 years, according to a report by Peru’s National Meteorology and Hydrology
Service (SENAMHI); see Vergara 2008.
8. The number of glaciers that have entirely melted is small; many are reduced
to small remnants. However, these small remnants can persist, especially those
at higher elevations that face away from the sun (southern exposures in the
Southern Hemisphere) (Ben Orlove, pers. comm.; IPCC 2007b).
9. Referring back to the rural poverty discussion, it should be noted that there
are very few small-scale poor farmers in northern Mexico. They are mostly
located in the southern part of the country.
Water Scarcity, Climate Change, and the Poor 55
criteria for water allocation have been developed. But as Tortajada points out,
this innovative initiative is not enough to affect water management in the
basin overall. The COTAS are not managed in a uniform manner, and they
exist in only two of the basin’s five states.
20. This exercise was undertaken by Mendelsohn (2008) in a background paper
for De la Torre, Fajnzylber, and Nash (2009).
21. According to Amat y Leon and his coauthors (2008), climate change and vari-
ability losses could amount to US$30 billion annually in the four Andean
countries by 2025.
22. Efficient water pricing would allocate water to its highest-valued uses. It
should reduce wasteful use, which is a big problem particularly affecting agri-
culture, the biggest consumer of water. However, in the absence of programs
ensuring the poor a basic income through a cash safety net, which they could
spend as they wish, paying the same price for each good as everyone else, it
may be necessary to introduce a measure to protect the poor, who otherwise
might not be able to afford water for their basic needs. For example, a life-
line tariff would price a quantity of water sufficient for basic needs very low,
with a higher price applying to quantities over the basic needs threshold. The
drawback of this option is that it requires metering, which rarely is present.
Another option could be a system of fixed charges based on neighborhood,
with a very low (or no) charge in poorer neighborhoods.
23. This and the following paragraph are based on information from Wikipedia.org,
“Water Resources Management in Bolivia.”
24. More details on this index, along with additional indexes, are provided in
Gleick and others 2002.
25. More information can be found at http://ocwr.ouce.ox.ac.uk/research/
wmpg/cvi.
26. Thanks to R. Cessti for bringing our attention to this recommendation.
References
Alfaro, J. F., and J. Marin. 1994. “On-farm Water and Energy Use for Irrigation in
Latin America.” In H. Garduño and F. Arreguín-Cortés (eds.), Efficient Water
Use. Montevideo, Uruguay: UNESCO–ROSTLAC.
Amat y Leon, Carlos, B. Seminario, M. P. Cigaran, S. Bambaren, L. Macera, M. T.
Cigaran, and D. Vasquez. 2008. El Cambio Climatico no Tiene Fronteras—
Impacto del Cambio Cliamtico en la Comunidad Andina. Secretaria General
De la Comunidad Andina. Lima, Peru.
Anton, D. J. 1993. Thirsty Cities: Urban Environments and Water Supply in Latin
America. Ottawa, Canada: International Development Research Center.
Water Scarcity, Climate Change, and the Poor 57
Arnell, N. W. 2004. “Climate Change and Global Water Resources: SRES Scenarios
and Socio-economic Scenarios,” Global Environmental Change 14: 31–52.
Avissar, R., and D. Werth. 2005. “Global Hydroclimatological Teleconnections
Resulting from Tropical Deforestation,” Journal of Hydrometeorology 6: 134–45.
Bolivian National Statistical Institute. 2009. http://www.ine.gov.bo (accessed
May 21, 2009).
Bradley, R., M. Vuille, H. F. Diaz, and W. Vergara. 2006. “Threats to Water Supplies
in the Tropical Andes,” Science 312 (23), June.
Carvajal, Liliana. 2007. Impacts of Climate Change on Human Development.
Background paper to Fighting Climate Change: Human Solidarity in a Divided
World, Human Development Report 2007/08 (UNDP 2007/08). New York:
United Nations Development Program.
Coudrain, Anne, Bernard Francou, and Zbigniew Kundzewicz. 2005. “Glacial
Shrinkage in the Andes and Consequences for Water Resources—Editorial.”
Hydrological Sciences–Journal des Sciences Hydrologiques 50 (6): 925–32.
Cox, P. M., R. A. Betts, M. Collins, P. P. Harris, C. Huntingford, and C. D. Jones.
2004. “Amazonian Forest Dieback under Climate Carbon Cycle Projections
for the 21st Century.” Theoretical and Applied Climatology 78: 137–56.
DANIDA (Danish International Development Agency). 2008. “Dialogue on
Climate Change Adaptation for Land and Water Management.” Draft
Concept Paper. Copenhagen: DANIDA.
De la Torre, Augusto, Pablo Fajnzylber, and John Nash. 2009. Low Carbon, High
Growth. Latin American Responses to Climate Change. An Overview. World
Bank Latin American and Caribbean Studies. Washington, DC.: World
Bank
Engle, N. L. 2007. “Adaptive Capacity of Water Management to Climate Change
in Brazil: A Case Study Analysis of the Baixo Jaguaribe and Pirapama River
Basins.” Master’s thesis, School of Natural Resources and Environment,
University of Michigan. http://hdl.handle.net/2027.42/50490.
Estrada, Daniela. 2006. “Latin America: Family Farms—Durable but Fragile.” Inter
Press Service (IPS) News Agency. http://ipsnews.net/news.asp?idnews=34996.
Fay, Marianna, ed. 2005. The Urban Poor in Latin America. World Bank Report.
Washington, DC: World Bank.
Fernandez, B. P., and L. B. Graham. 1999. “Sustainable Economic Development
through Integrated Water Resources Management in the Caribbean.” Paper
Presented at the II Water Meeting, Montevideo, Uruguay, June 15–18.
Formiga-Johnsson, R. M., and K. E. Kemper. 2005. Institutional and Policy Analysis
of River Basin Management—The Jaguaribe River Basin, Ceará, Brazil. Policy
Research Working Paper 3649, World Bank, Washington, DC.
58 Rossing
Maskrey, A., Gabriella Buescher, Pascal Peduzzi, and Carolin Schaerpf. 2007.
Disaster Risk Reduction: 2007 Global Review. Consultation ed. Prepared for the
Global Platform for Disaster Risk Reduction First Session, Geneva, June 5–7.
Mendelsohn, R. 2008. “Impact of Climate Change on the Rio Bravo River.”
Background paper for De la Torre, Fajnzylber, and Nash, 2009. World Bank
Latin American and Caribbean Studies, World Bank, Washington, DC.
Miller, K. 2006. Land under Siege. Recent Variations in Sea Level through the
Americas. Faculty of Engineering, University of the West Indies, St. Augustine,
Trinidad and Tobago, West Indies.
Milly, P. C. D., K. A. Dunne, and A. V. Vecchia. 2005. “Global Pattern of Trends in
Streamflow and Water Availability in a Changing Climate.” Nature 438:
347–50.
National Environmental Advisory Board/Ministry of Health and the Environment.
2000. “Initial National Communication on Climate Change. St. Vincent and
the Grenadines.” General Secretariat, Organization of American States,
Washington. DC.
National Research Council. 2003. Cities Transformed: Demographic Change and Its
Implications in the Developing World. Panel on Urban Population Dynamics,
Committee on Population, Division of Behavioral and Social Sciences and
Education. Washington, DC: National Academies Press.
Nohara, D., A. Kitoh, M. Hosaka, and T. Oki. 2006. “Impact of Climate Change
on River Runoff.” Journal of Hydrometeorology 7: 1076–89.
OECD (Organization for Economic Cooperation and Development). 2003.
OECD Territorial Reviews: Mexico. Paris: OECD.
———. 2007. Ranking Port Cities with High Exposure and Vulnerability to Climate
Extremes: Exposure Estimates. December. OECD Environment Directorate.
Paris: OECD.
Ohno, A, A. Marui, E. S. Castro, A. A. Reyes, D. Elio-Calvo, H. Kasitani, Y. Ishii,
and K. Yamaguchi. 1997. “Enteropathogenic Bacteria in the La Paz River of
Bolivia.” American Journal of Tropical Medical Hygiene 57(4): 438–44.
Oki, T., and S. Kanae. 2006. “Global Hydrological Cycles and World Water
Resources.” Science 313: 1068–72.
Porto, M. 1998. “The Brazilian Water Law: A New Level of Participation and
Decision Making.” International Journal of Water Resources Development 14 (2):
175–82.
Quijandria, B., A. Monares, and R. U. de Peña Montenegro. 2001. Assessment of
Rural Poverty, Latin America and the Caribbean. Santiago, Chile: International
Fund for Agricultural Development.
Ravallion, Martin, Shaohua Chen, and Prem Sangraula. New Evidence on the
Urbanization of Global Poverty. Background paper for World Development
60 Rossing
U.S. Army Corps of Engineers. 2004. Water Resources Assessment of the Bahamas.
Mobile District and Topographic Engineering Center, United States
Southern Command. http://www.sam.usace.army.mil/en/wra/Bahamas/
BAHAMASWRA.pdf.
Vergara, W. 2008. “Retracting Glacier Impacts: Economic Outlook in the Tropical
Andes.” Latin America and Caribbean Department, April 23, World Bank,
Washington, DC.
Vergara, W., A. M. Deeb, A. M. Valencia, R. S. Bradley, B. Francou, A. Zarzar,
A. Grünwaldt, and S. M. Haeussling. 2007. “Economic Impacts of Rapid
Glacier Retreat in the Andes.” Eos 88 (25): 261–64.
WBGU (German Advisory Council on Global Change). 2008. “Climate Change
as a Security Risk.” London: Earthscan. http://www.wbgu.de/wbgu_jg2007
_engl.html.
Wester, P., D. J. Merrey, and M. de Lange. 2003. “Boundaries of Consent:
Stakeholder Representation in River Basin Management in Mexico and South
Africa.” World Development 31 (5): 797–812.
WMO/IADB (World Meteorological Organization and Inter-American
Development Bank). 1996. Water Resources Assessment and Management
Strategies in Latin America and the Caribbean. Proceedings of the
WMO/IADB Conference, San José, Costa Rica, May 6–11.
World Bank. 2000. “Argentina: Poor People in a Rich Country.” Report No. 19992-
AR, World Bank, Washington, DC.
———. 2001. “Attacking Brazil’s Poverty: A Poverty Report with a Focus on
Urban Poverty Reduction Policies.” Report No. 20475-BR, World Bank,
Washington, DC.
———. 2002. “Mexico Urban Development: A Contribution to a National Urban
Strategy.” Report No. 22525, World Bank, Washington, DC.
———. 2005. World Development Indicators 2005. Washington, DC: World Bank.
———. 2006. “Project Document for Regional Implementation of Adapation
Measures in Coastal Zones (SPACC) Project.” World Bank Latin America and
Caribbean (LAC)/Global Environment Facility, World Bank, Washington, DC.
———. 2008a. World Development Report: Agriculture for Development.
Washington, DC: World Bank.
———. 2008b. “Poverty Data: A Supplement to World Development Indicators
2008.” Washington, DC: World Bank.
———. 2008c. Equity and Efficiency in the Greenhouse: Responding to Climate
Change in Latin America and the Caribbean Volume I: Overview. Sustainable
Development Department (LCSSD) and Chief Economist Office
(LCRCE), Latin American and the Caribbean Region. Washington, DC:
World Bank.
62 Rossing
Latin America and the Caribbean are among the regions of the world most
prone to climate-related hazards. Flooding and landslides are common in
the region, with its complex river basin systems and mountainous terrain.
Tropical storms and hurricanes, formed in both the Pacific and Atlantic
Oceans, are frequent throughout, especially in the Caribbean hurricane
belt, which includes Central America. Climatic variability, in the form of
severe droughts, floods, and high winds throughout the hemisphere, has
been exacerbated by the recurrent El Niño phenomenon (IADB 2000).
It is therefore alarming that growing evidence suggests that climate
change and variability will increase the region’s exposure to natural haz-
ards and climate-related disasters (IPCC 2007b; Mendelsohn and Williams
2004; Nagy and others 2006; Vergara and others 2007). The long-term
effects of global warming include changes in average temperatures and
precipitation, deglaciation, and sea level rise. Climatic variability is also on
the rise; its effects are already noticeable. Many natural hazards are pro-
jected to become more intense and more common, and increased variabil-
ity will also create more surprises, such as natural disasters occurring in
succession or in places where they have never been experienced before.
Though most of the projections of the Intergovernmental Panel on
Climate Change (IPCC) are on a global or regional scale, the impact of
63
64 Rossing and Rubin
America and the Caribbean, storms that develop along the intertropical
convergence zone and the subtropical high-pressure zone dominate the
weather. As a result of these patterns, flooding tends to be associated
with hurricanes and other tropical storms, which generate very heavy rain-
fall continuing over several days (Charveriat 2000). In South America,
a major factor in the occurrence of floods is the El Niño–Southern
Oscillation (ENSO; see Box 3.1).2
Box 3.1
Workers lay sandbags along the Rio Grande, some 50 kilometers (31 miles) east
of Santa Cruz, January 25, 2008. The Red Cross reported that heavy rains and
flooding in much of Bolivia killed 30 people and caused hardship for nearly
25,000 families.
According to the Bolivian Red Cross, shortly after the disaster struck, Bolivia
had a preliminary response plan in place to provide humanitarian assistance to
a total of 1,200 families—600 in Santa Cruz, 300 in Potosí, and another 300 in
La Paz—in the form of food parcels and hygiene kits. “One of the most serious
problems that we encounter is that the secondary roads often are impassable
as a result of the floods, which makes it difficult to get supplies and personnel
specialized in assessment and response through to where they are needed,”
said Rubén Romero, disaster management delegate.
Source: Red Cross (2008), http://www.ifrc.org/docs/news/08/08013103/; and authors’ elaboration based
on fieldwork in Bolivia, in March 2008.
Climate Change, Disaster Hot Spots, and Asset Erosion 67
Figure 3.1 Climate Impacts of El Niño Phenomenon in Latin America and the
Caribbean, 1998
Severe
droughts
Excessive
rainfall
Drought Scarce
episodes precipitation
soil moisture deficit
+0.5+2.5°C
Increase in surface ocean temperatures +0.5+2.5°C
during the 1998 El Niño event +2.5+5.0°C
Severe droughts, Scarce
increased forest fires precipitation
in rainy
season
Intense rains High
in summer, temperature
dryness in the episodes
Andean region
Plentiful rains
in spring and
summer
IBRD 37782
MAY 2010
Sources: Adapted from IPCC 2001. FAO 2002. UNEP 2003. Climate impacts of El Niño Phenomenon in Latin America
and the Caribbean (2005). In UNEP/GRID-Arendal Maps and Graphics Library.
68 Rossing and Rubin
IPCC notes that changes in these types of extreme events due to climate
change and variability can already be observed.
As regards future small-scale atmospheric phenomena, such as hurri-
canes, cyclones, and tidal waves, the IPCC projections are much more
uncertain. The traditional IPCC models project a consistent increase in
precipitation intensity in future storms. That projection is supported by
more recent, higher-resolution models that project increases in peak wind
intensities with some consistency and a rise in mean and peak precipita-
tion intensities in future tropical cyclones with even greater consistency
(IPCC 2007a). Some of these disaggregated models have even projected
fewer but more intense storms, due to climate change and variability,
with potential to cause much more damage than a greater number of
less-intense storms.
Patterns of actual and predicted natural hazards show that the areas
most prone to feel the effects of climate change are also the poorest. The
areas at highest risk from storms and flooding include most of the
Caribbean (especially Cuba, Dominica, the Dominican Republic, Haiti,
Puerto Rico, and St. Lucia); Guatemala, Honduras, and Nicaragua in
Central America; and the Bolivian, Brazilian, and Peruvian Amazon
regions in South America. Areas highly vulnerable to drought include
northeast Brazil, the arid Andean region (particularly Bolivia and Peru),
and the north of Mexico.
Table 3.3 Weather-Related Disaster Exposure Indicators in Latin America and the
Caribbean, by Subregion, 1970–2008
South America Caribbean Central America
Frequency per square kilometer (%) 0.3 13.4 6.2
Fatalities per million inhabitants 122.0 323.1 1,043.6
Population affected per thousand
inhabitants 249.2 665.3 524.3
Source: Authors’ elaboration based on Wikipedia.org and United Nations Population Division 2006.
Notes: Frequency per square kilometer is calculated based on cumulative events between 1970 and 2008,
divided by total square kilometers. Deaths per million inhabitants are calculated by dividing cumulative deaths
1970–2008 by 2005 population estimates. Affected population per thousand inhabitants is calculated by dividing
the cumulative affected population 1970–2008 by 2005 population estimates. From a social perspective,
calculating damages as a percentage of gross national product (GDP) is a highly imperfect measure for the
severity of the disaster. For most vulnerable people, the impact on GDP from a disaster will in most cases be neg-
ligible. That is why much higher damage occurs in developed nations (even relative to GDP) in comparison to
developing nations.
Central America appears the most vulnerable region with 1,043 fatalities
per million population, followed by the Caribbean (323 fatalities per
million) and South America (122 fatalities per million). The Caribbean is
by far the hardest-hit region when land size is taken into account. Thus,
although weather-related disasters are more frequent in South America,
they cause more fatalities per capita in Central America. One reason may
be that relatively more infrastructure is in place in South America than in
Central America.
Individual countries’ vulnerability relates closely to their size, gover-
nance, and economic wealth. Brazil—the largest country in South
America—had by far the largest number of disasters between 1970 and
72 Rossing and Rubin
2008, totaling 2.7 a year, but it experienced among the lowest rates of fatal-
ities per capita. Peru and Colombia experienced moderately fewer disasters
but suffered significantly greater damage and fatalities. In the Caribbean,
Cuba and Haiti ranked the highest in disaster frequency, but Cuba had only
3 percent of the fatalities that Haiti experienced. Dominica, St. Lucia, the
Dominican Republic, and Puerto Rico had the next-highest loss of lives. In
Central America, Mexico had many more disasters than other countries but
relatively few fatalities. Guatemala, Nicaragua, and Honduras, with moder-
ate numbers of disasters, ranked highest in fatalities per capita. Belize and
Panama have the best record in Central America, with low numbers of dis-
asters and fatalities in both absolute and relative terms (Charveriat 2000)
(for an example, see the account of flooding in Belize in chapter 10).
Mortality rates are an important indicator of the erosion of human
assets from climate-related disasters, but the number of survivors affected
is at least as significant. EM-DAT defines “affected populations” as those
requiring immediate assistance during a period of emergency to obtain
basic survival needs such as food, water, shelter, sanitation, and immedi-
ate medical assistance. As table 3.4 shows, for each person who died in a
climate-related calamity in the LAC region during 1970–2008, another
1,600 were adversely affected to the extent that they required assistance
to survive. The table also shows that whereas droughts led to a relatively
small loss of life, they affected more than 60 million people between
1970 and 2008, while floods affected 51 million, and storms 31 million.5
In other words, in those locations, the hazards only rarely become disasters.
Paradoxically, flooding is less common in the parts of the region that have
the most deaths from flooding: the northern parts of South America and
Mexico and extensive parts of Central America and the Caribbean
islands. The explanation lies in predictability. Flooding occurs regularly
along the Amazon, and the indigenous peoples living nearby have
adjusted their way of living to its seasonal occurrence. Only when flood-
ing is higher than usual—or fails to materialize—does it cause problems.
People living in areas that do not regularly experience floods have no
social memory of them, and what to do in response, and are therefore
more likely to be caught by surprise. Then such hazards can turn into
disasters in which people are killed or suffer injury or material losses.
Similarly, while the most frequent storm hazards arise from the
Pacific Ocean along the coast of Mexico, the highest mortality rates from
storms that became disasters are caused by the storms in the Atlantic
Ocean, which strike the Caribbean islands as well as Honduras and
Nicaragua. Although weather-related hazards are more frequent in South
America, they cause proportionately more fatalities in Central America,
and particularly in the Caribbean, simply because a larger proportion of
the population there is exposed to them. The data comparisons also
emphasize that disaster occurrence and impact are related to the level of
socioeconomic development: poorer countries and areas tend to suffer
proportionately more fatalities and injuries.
people from adverse effects of climate change should try to support exist-
ing coping strategies so as to strengthen a community’s resilience rather
than focus solely on post-disaster recovery measures.
In addition to the need to focus efforts on risk reduction and the need
to be proactive rather than just reactive, the question must be asked: Are
there socioeconomic determinants of natural disasters? This section dis-
cusses how climate-related disasters can affect the asset base of the poor
in the LAC region, to compound their existing vulnerability by seriously
constricting and often destroying their livelihoods.
food and income, coupled with lack of physical and financial adaptive
capacity, means that poor farmers are also at increased risk of harm from
slow-onset disasters.
The majority of low-income settlements in both urban and rural areas
in the LAC region tend to be in inhospitable areas prone to flooding,
landslides, or drought. Such low-income settlements, at constant risk from
hurricanes and flooding, can be found in many of the Caribbean and
Central American countries. For example, in St. Vincent many households
in Georgetown and on the outskirts of Kingston worry that their homes
will crumble during the next storm in consequence of their perilous loca-
tion on steep slopes. And in Belize City, people live in constant fear that
the next storm that passes through will bring severe flooding, as occurred
during Hurricane Hatti, when water levels rose to 10–12 feet in some of
the city’s poorest districts.6 The poor in urban settlements in the region’s
cities and megacities have their livelihoods destroyed year after year
because they reside in areas prone to mudslides and flooding during the
rainy season.
Poor people are constrained in their choice of settlement by property
rules and landowners. They may choose hazardous areas voluntarily, if
those areas seem to improve their access to resources or increase their
income-generating possibilities. Hence, environmental hazard might be
outweighed by the perceived benefits of living in such hazard-prone
areas as fertile volcanic slopes, plains nourished by flood alluvia, river-
banks, or other areas shunned by formal settlements because of their
high exposure to risk but which nonetheless offer access to cities and
hence income-earning opportunities (Rapp 1991, as cited by Main and
Williams 1994). In rural areas, poverty is also a direct cause of environmen-
tal degradation (chapter 4), as the extremely poor—who constitute a high
percentage of the rural poor—are often forced to deplete their surrounding
natural resources for survival (Echeverria 1998). A strong connection also
exists between poor-quality housing and low income, especially where low-
income people cannot get mortgage financing. Lack of land ownership and
lack of rental possibilities exacerbate these links.
It is typically the poor who bear the brunt of weather shocks and who
recover most slowly. For example, when Hurricanes Wilma and Stan
affected the Yucatan peninsula and Chiapas, in Mexico, in 2005, they
caused relatively greater damage to the assets of the poor than to those of
the nonpoor. More than 50 percent of the losses from Hurricane Wilma in
Yucatan (mostly luxury properties in Cancun) were insured (Carpenter
2006). By contrast, almost all the losses from Hurricane Stan were
Climate Change, Disaster Hot Spots, and Asset Erosion 77
Natural assets. Natural assets comprise the natural resource stock from
which households derive resources useful to their livelihood. It is beyond
the scope of this chapter to analyze the intricate causal relationship
between weather-related natural disasters and natural resources per se.9
But box 3.2 illustrates how climate change and variability will compound
existing stresses on vital natural assets.
Climate Change, Disaster Hot Spots, and Asset Erosion 79
Box 3.2
Box 3.3
(continued)
82 Rossing and Rubin
asked the hypothetical question: “If you could ask for three things, what would they
be?” Invariably a request for helping those hardest hit led the list, even when those
people were not represented in the group being interviewed. Hence, even though
some people essentially lost their livelihood, the strong bonding social capital in
the community enabled them to remain in the village, rather than be forced to mi-
grate to La Paz to find other income opportunities.
Another subsistence farmer community, Palca, was also severely affected by
the floods but appeared to react differently. Floods destroyed a century-old sys-
tem of irrigation canals dug into the mountainside. As a result, the community was
cut off from its crucial water supply, which is fed by glacial melt above them. To
restore the water supply and return the farms to their former productive capacity,
the infrastructure would need repair, involving installation of new plastic pipes
worth US$10,000, a project that could have been implemented quite quickly. How-
ever, the farmers appeared very despondent and did not think it would be possi-
ble to mobilize the assistance and funds for the repair. Although they were collec-
tively dependent on the irrigation system, they were all acting as individuals. They
also lacked the level of organization to jointly seek assistance or loans to finance
the replacement of the pipes. The community therefore appeared to lack suffi-
cient bonding social capital, as well as linking social capital to its municipality
government. Instead the coping mechanism that many households adopted was
to send the young to La Paz in search of work, so that they could send money
home to the family members left behind. Also, small children were taken out of
school to work in the fields alongside the elderly in the community.
It appeared that because Palca was so destitute, with so few assets and options
available, their supply of social capital was quickly exhausted. In contrast, the com-
munity in Mecapaca, situated at a lower altitude and in flatter terrain, could grow
a variety of crops and therefore had more diversified income opportunities. Never-
theless, that community was also very conscious of the threat to its sustainability if
its young people left for La Paz, and it put much effort into providing opportunities
for young people locally as incentives for them to remain in the community.
Source: Authors.
to quality housing and land usually translates into less capacity to with-
stand damage. It also means that authorities often do not provide the
services and infrastructure needed to cope with climate hazards.
Where housing and land are permanently affected by climate hazards,
it becomes very difficult for poor households to improve their lives and
reduce their long-term vulnerability. In Honduras, when Hurricane Mitch
destroyed people’s homes, they lost their shelter and their tools and mate-
rials for making a living. Many farmers in one ladino village in the
Noroiente Region of the country also lost fertile land, as flooding washed
away the productive topsoil and deposited debris and rocks all over the
fields (De la Fuente 2007). The same happened to the Huayhuasi com-
munity in Mecapaca, Bolivia (box 3.3).
Policy Perspectives
Climate-induced changes in resource flows—especially through erosion
of key assets—will affect the viability of some livelihoods unless measures
are taken to protect and diversify them through adaptation. Not all
groups are affected equally, even within a given natural disaster hot spot,
however. Hence there is a rationale for applying a context-specific vulner-
ability analysis using an asset-based framework to ensure a strong match
between the needs of particular vulnerable groups and whatever inter-
ventions are planned to assist them.
As part of such asset-based vulnerability analyses, it may be helpful to
distinguish between the natural hazards caused by increased climate vari-
ability of the type we can already observe and the projected long-term
trends (30 years and up) for climate change, which might cause more
severe natural hazards. Such a distinction has implications for the analy-
sis of asset erosion and vulnerability, particularly with respect to notions
of resilience, adaptation, and development. The distinction, in turn, will
affect how hazard-risk management is considered for and by the poorest
groups. In other words, is it more important to manage the climate vari-
ability that is already being felt, or to concentrate on the long term, to
increase resilience and prepare people to adapt to the natural hazards to
come 30 years from now?
Governance structures are pivotal in determining the impact of natu-
ral hazards on livelihoods. The most vulnerable groups often suffer dis-
proportionately from weak, unresponsive local institutions. Community
participation, voice coalition, and local governance relating to natural
Climate Change, Disaster Hot Spots, and Asset Erosion 87
Notes
1. Emergency Events Database (EM-DAT), http://www.emdat.be.
2. The periodic El Niño events are a natural phenomenon that has occurred for
centuries. Ocean and atmospheric conditions in the Pacific tend to fluctuate
between El Niño (warming) and La Niña (cooling). The fluctuations are semi-
regular and tend to appear every three to six years. A more intense phase of
either event may last for about a year.
3. Some of this increase is likely to be due to improved reporting, and some can
be attributed to population movements and development—more people live
in exposed locations than before, and more material damage occurs because
of development where previously there was none or very little.
4. This finding is based either directly on data from the Center for Research on
the Epidemiology of Disasters (CRED) or on simple tabulations thereof.
5. These outcomes derive from the nature of the various phenomena in the
region (for instance, droughts affecting parts of Brazil) and their respective
rates of occurrence. The fact that windstorms affected fewer people is linked
to the fact that these phenomena generally affect smaller geographical areas
than do floods or droughts.
6. Information obtained during fieldwork in Belize and St. Vincent in 2008.
7. Thanks to Nicolas Perrin of the World Bank for drawing our attention to the
need for including the aspect of governance at this stage. These aspects are also
addressed in great detail by the World Bank project, led by Nicolas Perrin and
Eija Pehu, supporting area-based development initiatives to enhance adaptive
capacity and maintain the resilience of local actors and institutions to climate
change, and by a research initiative put forward by Arun Agrawal and Nicolas
Perrin to shed light on the role of social capital in addressing climate hazards.
8. An early-warning system could have prevented the deaths of several Miskito
fishermen, as well as brought women and children to safety before the hurri-
cane hit (Vianica 2007; MADRE 2007; Rodriguez 2007).
88 Rossing and Rubin
9. Thorough analyses have already been undertaken by others (such as the earlier
IPCC reports and the Millennium Ecosystem Assessment).
10. Observation made on field visit to St. Vincent, June 2008.
11. A significant discrepancy appears between these figures and the official esti-
mates of unemployment from November 1999, which registered 89,000
unemployed, or only 3.7 percent, compared with 3 percent before Mitch
(Boletin del Comisionado 1999). One explanation for this discrepancy could
be that much of the frictional unemployment was absorbed between March
and November 1999.
12. A survey of subsistence farmers in Chiapas, Mexico, indicated that the respon-
dents tended to depend more on relatives living in the same community and
aid from the government than on financial instruments such as insurance and
credit. Hence the challenge of integrating financial prevention measures into
the region’s adaptive strategy.
References
Adger, W. Neil. 1999. “Social Vulnerability to Climate Change and Extremes in
Coastal Vietnam.” World Development 27 (2): 249–69.
———. 2003. “Social Capital, Collective Action, and Adaptation to Climate
Change.” Economic Geography 79 (4): 387–404.
———. 2006. “Vulnerability,” Global Environmental Change 16: 268–81.
Anderson, M. G., and L. Holcombe. 2008. “Community-Based Landslide Risk
Reduction: Proof of Concept.” Presentation at the World Bank, Washington
DC, June 16.
Boletin del Comisionado, La Prensa, November 16, 1999.
Bouma, M. J., S. R. Kovats, S. A. Goubet, H. St. J. Cox, and A. Haines. 1997.
“Global Assessment of El Nino’s Disaster Burden.” Lancet 1435 (8): 350.
Carpenter, G. 2006. “Tropical Cyclone Review 2005.” Instrat Briefing, January. http
://gcportal.guycarp.com/portal/extranet/popup/pdf/GCBriefings/Tropical_
Cyclone_Review_2005.pdf.
Carvajal, Liliana. 2007. “Impacts of Climate Change on Human Development.”
Background Paper to UNDP 2007, Fighting Climate Change: Human Solidarity
in a Divided World, Human Development Report 2007/08. New York: Oxford
University Press.
Charveriat, C. 2000. Natural Disasters in Latin America and the Caribbean: An
Overview of Risk. Working Paper 434, Research Department, Inter-American
Development Bank, Washington, DC.
De la Fuente, A. 2007. Climate Shocks and their Impacts on Assets. Occasional paper
for UNDP 2007, Fighting Climate Change: Human Solidarity in a Divided World,
Human Development Report 2007/08. New York: Oxford University Press.
Climate Change, Disaster Hot Spots, and Asset Erosion 89
Echeverría, R.G. 1998. “Rural Poverty Reduction.” Bank Strategy Paper, Inter-
American Development Bank, Washington, DC.
Hodell, D. A., J. H. Curtis, and M. Brenner. 1995. “Possible Role of Climate in
the Collapse of Classic Maya Civilization.” http://www.ncdc.noaa.gov/paleo/
drought/drght_mayan.html.
IADB (Inter-American Development Bank). 2000. Development beyond
Economics: Economic and Social Progress in Latin America, 2000 Report.
Washington, DC: Inter-American Development Bank.
IBGE (Instituto Brasileiro de Geografia e Estatistica). 2004. “IBGE Releases
Previously Unseen Profile of Food Security in Brazil.” http://www.ibge.gov
.br/english/presidencia/noticias/noticia_visualiza.php?id_noticia=600&id
_pagina=1.
IPCC (Intergovernmental Panel on Climate Change). 2007a. Climate Change
2007: The Physical Science Basis. Contribution of Working Group I to the
Fourth Assessment Report of the IPCC. Geneva: IPCC.
———. 2007b. Climate Change 2007: Impacts, Adaptation, and Vulnerability.
Contribution of Working Group II to the Fourth Assessment Report of the
IPCC. Geneva: IPCC.
Lemos, M. C., D. Nelson, T. Finan, and R. Fox. 2004. “The Social and Policy
Implications of Seasonal Forecasting: A Case Study of Ceara, Northeast
Brazil.” NOAA Report, University of Arizona.
MADRE News. 2007. “Hurricane Felix Tears through Nicaragua; Local
Communities Destroyed.” September 5. http://www.madre.org/index/press-
room-4/news/hurricane-felix-tears-through-nicaragua-local-communities-
destroyed-16.html.
Main, H., and S. W. Williams, eds. 1994. Environment and Housing in Third World
Cities. New York: John Wiley and Sons.
McMichael, A.J., R. E. Woodruff, and S. Hales. 2003. “Climate Change and
Human Health: Present and Future Risks.” Lancet 859 (69): 367.
Mendelsohn, Robert, and Larry Williams. 2004. “Comparing Forecasts of the
Global Impacts of Climate Change,” Mitigation and Adaptation Strategies for
Global Change 9: 315–33.
Moser, C.O. 1996. Confronting Crisis—A Comparative Study of Households.
Environmentally Sustainable Development Studies and Monograph Series
No. 8. Washington, DC: World Bank.
Nagy, Caffera, Aparicio, et al. 2006. “Understanding the Potential Impact of
Climate Change and Variability in Latin America and the Caribbean.” Report
prepared for the Stern Review on the Economics of Climate Change.
http://www.hm-treasury.gov.uk/media/6/7/Nagy.pdf.
Narayan, D. 1997. Voices of the Poor: Poverty and Social Capital in Tanzania.
Washington, DC: World Bank.
90 Rossing and Rubin
of the Earth Simulator” Latin America and the Caribbean Region, Sustainable
Development Department, World Bank, Washington, DC.
Vianica. Hurricane Felix Information Center. 2007. “Hurricane Felix and the
Miskito Cays.” http://www.vianica.com/projects/1/
Vos, R., M. Velasco, and E. De Labastida. 1999. Economic and Social Effects of El
Niño in Ecuador. Washington, DC: Inter-American Development Bank.
WMO (World Meteorological Organization). 1999. The 1997–1998 El Niño
Event: A Scientific and Technical Retrospective. No. 905. Geneva: World
Meteorological Organization.
World Bank. 2000–01. World Development Report 2000/2001: Attacking Poverty.
Washington, DC: World Bank. http://go.worldbank.org/L8RGH3WLI0.
———. 2005. The Urban Poor in Latin America. Ed. Marianne Fay. Directions in
Development. Washington, DC: World Bank.
World Bank. 2008a. “At a Glance: Poverty Numbers in LAC.” http://www.worldbank
.org/lacpoverty
World Bank. 2008b. Equity and Efficiency in the Greenhouse: Responding to Climate
Change in Latin America and the Caribbean Volume I: Overview. Sustainable
Development Department (LCSSD) and Chief Economist Office (LCRCE),
Latin American and the Caribbean Region. Washington, DC: World Bank.
World Bank. 2008c. “Poverty Data: A Supplement to World Development
Indicators 2008.” Washington, DC: World Bank.
World Bank. 2008d. World Development Report: Agriculture for Development.
Washington, DC: World Bank.
WWF (World Wildlife Fund). 2006. Up in Smoke? Latin America and the
Caribbean—The Threat from Climate Change to the Environment and Human
Development. Third Report from the Working Group on Climate Change and
Development. London: New Economics Foundation.
Zapata-Martí, Ricardo. 2006. Los efectos de los desastres en 2004 y 2005: La necesi-
dad de adaptación de largo plazo. Punto Focal de Evaluación de Desastres,
ECLAC, Mexico City.
Zhao, Y., C. Wang, S. Wang, and L. V. Tibig. 2005. “Impacts of Present and Future
Climate Variability on Agriculture and Forestry in the Humid and Sub-humid
Tropics,” Climatic Change 70: 73–116.
CHAPTER 4
Agrarian Livelihoods
and Climate Change
Jørgen E. Olesen
93
94 Olesen
Box 4.1
region’s population, are indigenous, and most of them live in rural areas.
Most rural indigenous families live in extreme poverty, generally having
little or no schooling, few or no productive resources, limited knowledge
of production, few work skills, little or no access to basic and rural pro-
duction resources, and little or no political voice.
Small farmers make up 27 percent of the rural poor. They possess
(with or without title) small plots of land in arid or semiarid regions, on
hillsides, or on the fringes of irrigated valleys. They farm on ecologically
fragile lands subject to climatic variation and uncertainty. In many cases,
they raise crops and livestock on hillsides, where their activities lead to
loss of natural vegetation and soil erosion. These small farmers combine
agricultural production with seasonal off-farm work. Men take responsi-
bility for preparing the land and for harvesting, while women and children
tend the livestock. When men migrate in search of seasonal employment,
the women take over all farm activities.
Subsistence or landless farmers account for 20 percent of the rural poor
and are among the poorest of rural people. This group only has access to
leased land, whether as tenants or sharecroppers. Families depend on
Agrarian Livelihoods and Climate Change 97
Sources of Vulnerability
Climate change and variability are but two of many sources of vulnera-
bility for the rural poor. Like others worldwide, the LAC region’s poor
rural households suffer social and economic exclusion, often linked to
ethnicity and gender; lack of access to basic services (health, education,
housing); and very low levels of income and other assets. Poor rural com-
munities are often geographically isolated, with unreliable systems of
communication and services. Many are untouched by conditional cash
transfer programs. To safeguard themselves, they create deeply rooted
forms of reciprocity to maintain social and family relationships, and they
depend on extended family networks, or social bonding capital (defined
in chapter 1), for social and economic survival. Their livelihoods depend
on fragile natural resources that in many cases are already being rapidly
degraded. Particularly in drylands and other marginal lands, climate
change is likely to speed up the degradation.
Particularly important processes that affect the natural resources used
by poor farmers and herders are the conversion of rangelands or silvo-
pastoral systems to arable lands and the shortening of fallow periods, lead-
ing to land degradation and overall lower productivity. Eventually the
degraded lands may become desert, incapable of supporting life. These
processes are often driven by increasing population pressure; the expan-
sion of commercial, export-oriented agriculture; or the introduction of
98 Olesen
IBRD 37783
MAY 2010
Deforestation
Excessive deforestation has been taking place in tropical areas of Latin
America and the Caribbean for the past several decades. Among its causes
are population growth, extensive logging, an expanding agricultural area,
and lagging agricultural yields. The forested area is declining in all LAC
countries except Uruguay, mostly at rates of 0.5 percent to 1.5 percent
annually (Velarde 2004). At the current rate, LAC forests are projected
to shrink by 9 percent between 2000 and 2020, and the areas under per-
manent pasture, permanent crops, and arable crops are projected to
increase by 2 percent, 25 percent, and 13 percent, respectively.
In many tropical countries, most of the deforestation results from the
actions of poor subsistence cultivators. In the Amazon, however, those
farmers contribute only about 30 percent of deforestation. The other
70 percent is caused by the encroachment of cattle ranches (Nepstad and
others 2008), largely as an indirect result of an expansion of soybean
farming. Most soybean cultivation takes place outside the rainforest in the
neighboring cerrado grassland ecosystem and in areas that have already
been cleared. But as cattle farms and the lands of some subsistence farm-
ers are converted to soybean cultivation, cattle and subsistence farmers
turn to forest clearing to obtain new land. Studies have also shown a
strong association between logging and future clearing for settlement and
farming.
farming operations and small rural enterprises were left with no technical
assistance and services.
Another trend that has been causing difficulties for small producers is
the vertical integration of food supplies. Over recent decades, a handful
of vertically integrated, transnational corporations have gained increasing
control over the global trade in food, its processing, and sale. The biggest
chains, most of them owned by multinational giants, now control 65 per-
cent to 95 percent of supermarket sales in Latin America and the
Caribbean (FAO 2007). While the expansion of supermarkets could pres-
ent an opportunity for small farmers to reach new markets, it also pres-
ents a substantial risk of greater marginalization and even deeper poverty
(Berdegue and Ravnborg 2007). For example, between 1997 and 2001,
more than 75,000 Brazilian dairy farmers were delisted by the 12 largest
milk processors during consolidation of the market. Most of them pre-
sumably went out of business. Similar structural change in agriculture is
happening elsewhere in the region. It need not lead to increased poverty,
if society can provide other and possibly better job opportunities and
sources of livelihood to replace income from agriculture.
Impacts on Agriculture
Climate change and variability affect cropping systems directly and indi-
rectly through a range of pathways (Tubiello, Soussana, and Howden
2007). The effects are mediated through the farmer’s management of the
interactions between crops and their growth environment, which depend
crucially on available resources, including climate, soil, water, nutrients,
genetic diversity, and machinery or labor.
The projected increase in greenhouse gases will affect agro-ecosys-
tems directly, mainly by increasing photosynthesis at higher atmospheric
concentrations of carbon dioxide (CO2), and indirectly by altering cli-
mate, with consequent effects on the functioning of ecosystems (Olesen
and Bindi 2002). Increasing temperature affects plant development.
With warming, the start of active growth is advanced and plants develop
faster, which reduces crop duration. An analysis of global production
shows that cereal crop yields are reduced with increasing temperatures
(Lobell and Field 2007). The effects of higher temperatures on yields
will be greatest in tropical crops such as coffee that in the present warm
tropics perform close to the optimum, so that warming will often lead to
relatively large yield reductions (box 4.2). However, extreme events
such as cyclones, droughts, and floods cause greater damage to commer-
cial crops than changes in mean climate variables alone (Porter and
Semenov 2005).
Simulations of crop yield changes under climate change show large
differences across the region (De la Torre, Fajnzylber, and Nash 2009).
Considering the direct effect of higher atmospheric concentrations of
CO2, models using the Intergovernmental Panel on Climate Change’s
(IPCC) A2 emission scenario suggest that by 2080, crop yield changes
will range from reductions of 20 percent to 30 percent in such countries
as Bolivia, Mexico, Peru, and Venezuela, to increases of up to 5 percent
in Argentina (Cline 2007; Parry and others 2004).2 For smallholders, a
mean reduction in maize yields can be expected by 2050 (Jones and
Thornton 2003). With a reduction in land suitable for growing coffee in
Brazil, and the stresses caused by higher temperatures and lower rainfall
for coffee producers in Mexico (Gay and others 2006; IPCC 2007b),
cultivation of coffee will shift to other parts of the region, such as
Argentina (box 4.2).
Climate variability and climate change affect livestock production in
two ways: indirectly, by influencing the productivity of grassland and
hence the quantity and quality of fodder, and directly by causing more
heat- and disease-related stress and death (Zhao and others 2005). In
Agrarian Livelihoods and Climate Change 103
Box 4.2
regions that are currently warm and dry, climate change will mostly affect
both those aspects negatively.
Reduced water availability will particularly affect grain crops and live-
stock production in Central America (Mexico, Costa Rica, and Panama),
the Andes, and parts of Argentina, Brazil, and Chile (box 4.3). In drier
104 Olesen
Box 4.3
Droughts in Mexico
Mexico has a long and varied experience with drought, from descriptions in early
historical chronicles to contemporary climatic data and disaster declarations
(Liverman 1990). More than 85 percent of the country is arid or semiarid, and inter-
annual rainfall varies widely. Biophysical factors are the main reasons for vulnerability
in the northern and north-central regions, where rainfall is most variable, and in the
highlands, where the timing of rainfall is critical. In these regions, vulnerability is
likely to increase as a result of deforestation, overgrazing, and the projected cli-
mate changes. Social reasons for vulnerability vary greatly among regions and
population groups. More than half of Mexico’s cropland is operated by ejidos, a
form of cooperative land tenure, and drought losses are generally considerably
higher on ejidos than on private farms, for several reasons. After the Mexican Rev-
olution, the ejidos mostly received the less-productive and drier land. In many
regions they do not have much irrigated land and have problems obtaining credit.
Two policy changes effectively diminished the supply of water for small and tradi-
tional communal farmers, who also have less access to credit: the 1992 reform of
the water law led to higher water prices (Wilder and Lankao 2006), and the North
American Free Trade Agreement (NAFTA) encouraged export-oriented farmers to
switch into more water-demanding crops (Liverman 2000).
Mexican communities have developed many traditional technologies for cop-
ing with drought, such as sophisticated irrigation systems, growing adapted
maize varieties, and relying on traditional famine foods (such as cactus, agave,
and mesquite) in drought years (Eakin 2000). More recent adaptation schemes
that rely on irrigation have shown themselves vulnerable to multiyear droughts,
though attempts to make irrigation more efficient have been beneficial. Experi-
ence from the 1996 drought in northern Mexico has shown wide variation in the
responses of individual farmers to drought, with farmers planting and then pray-
ing for rain or government relief, and others deciding to abandon their land and
migrating to find work in other regions of Mexico or the United States. Small land-
holders were disproportionally affected, especially those who were in debt or
who farmed or ranched more fragile land (Liverman and Rosenberg 1996).
Source: Liverman 1990; 2000; Liverman and Rosenberg 1996.
areas, which largely coincide with the areas showing risk of desertifi-
cation, climate change and variability will likely speed up the saliniza-
tion and desertification of agricultural lands, compounding the effects
of reduced rainfall.
Agrarian Livelihoods and Climate Change 105
Social Impacts
Climate change will affect the livelihood sources of farmers, agricultural
laborers, and other rural workers whose jobs depend on the health of agri-
culture (IAASTD 2009). Climate change also interacts with other pres-
sures on livelihoods, such as the continued deforestation that reduces
the livelihoods of many indigenous people and threatens biodiversity
resources.
As noted above, climate change will affect the suitability of locally
adapted crops, giving rise to cultivation of new crop species and cultivars.
Because many rural people obtain most of their food from local agricul-
ture and forestry, local diets will need to take stock of the new crops being
grown. In the Andes, for example, other crops may replace potatoes, and
in some of the drier and warmer regions maize may need to be replaced
with more drought-tolerant crops such as sorghum or millet. Such new
crops may be less favored for human consumption, and that may lead to
more imports of food from outside. The local agricultural products may
then be used instead for animal feed—which may support the livelihoods
of farmers but increase the poverty of other rural people.
In many parts of the region, the effects of climate change on ecosys-
tems have begun to compromise the traditional livelihoods and lifestyles
of indigenous peoples who depend on them (box 4.4).
Changes in biodiversity, resulting from climate change and exacerbated
by human pressures on natural ecosystems, will likely reduce the availabil-
ity of many plant species used in modern as well as traditional medicine.
106 Olesen
Box 4.4
which some communities rely on for their livelihoods. Expectations are for more
powerful hurricanes and a continuing upward trend for intense precipitation
followed by extended periods of drought resulting from higher temperatures.
Indigenous people interviewed in Nicaragua said that even now 20 years after
Hurricane Joan, they have yet to fully recover the abundance of forest resources
that existed before the hurricane (especially lumber and wildlife). The failure of
forest resources to recover was closely linked to the expansion of the agricultural
frontier that was facilitated by large tracts of forest being felled by the hurricane,
allowing easy access to indigenous territories. In contrast, after 5 years yields
were close to levels prior to the event for fishing and perennial crops such as fruit
trees; yields recovered similarly after just one year for rice, beans, banana, plan-
tain, cassava, and maize. Indigenous peoples affected by climate change and
variability include the Miskitu, the Sumu-Mayangna complex (Ulwas, Twaskas,
and Panamaskas), and the Rama.
In the lowland forests of the Amazon region, drought is exacerbating the
effects on indigenous peoples of the ongoing deforestation. The threats from
climate change compound the effects of other dangers such as expansion of
the agricultural frontier, advancing colonization, political unrest, and illegal
coca cultivation. The very great diversity of fish fauna in the Amazon provides an
important food resource for indigenous communities. The successful reproduc-
tion of fish is dependent on the great rainy season that causes the river to rise.
Normally, it rains through April and May, when the river floods the lower and
higher alluvial plateaus and inundates extensive areas of the forest. During this
time, most wild fruits, fed by the water, ripen and fall into the rising waters. They
attract fish, which come to the surface and disperse all along the flooded forest
floor to feed on the plentiful food and to lay their eggs. However, the river no
longer rises and falls as before. In 2005 in the northern region, the river flooded
and fish hatched normally, but then suddenly, the river receded before the fish
were sufficiently mature, killing them. On the Amazon River (southern regime),
flooding has not been sufficient since 1999, and that has directly affected fish
reproduction. Abnormal river levels also directly affect the reproduction of tur-
tles, which require a drop in the river level at precisely the time they are ready
to lay their eggs on the beaches emerging out of the retreating river.
Indigenous peoples possess a strong awareness of complex seasonal rhythms.
The annual succession of seasons is of utmost importance for indigenous peo-
ples. It orders the timing of the horticultural cycle and the ritual practices that
(continued)
108 Olesen
help prevent illnesses and promote human well-being, and it is crucial for the repro-
duction of wildlife. Its disruption causes food insecurity and undermines the array of
solutions provided by cultural institutions and authorities.
According to key people interviewed in different areas, the natural signs that
they now perceive are alarming. Seasons have become irregular, the once regular
flood and ebb of rivers are now out of synchrony with the fall of wild fruits, and
heat is increasing. A recurrent lament voiced during all research field visits was
that seasonal variation has become so unpredictable that the cultural adaptation
strategies developed to tackle the normal span of variation no longer provide the
necessary security.
Without secure livelihood strategies, communities tend to break down. Some
indigenous peoples are compelled to change their livelihood so dramatically that
they lose conditions vital for the development and reproduction of their culture.
When climatic conditions become impossible to predict, the elders and traditional
leaders—who are the experts within traditional knowledge systems—lose credi-
bility. When traditional authorities cannot guarantee abundance and prosperity,
their status falls, and people look elsewhere for solutions to their problems, both
by turning to other bodies of knowledge and by migrating. This has been spoken
of among several peoples as leading to the end of life as indigenous peoples.
Not only do individual family members leave, but whole families are now
uprooting. Some communities are left ghostlike; in others, only children and the
elderly remain. This phenomenon is seen in communities across the region from
Argentina to Mexico. An example is the response to flood damage to irrigation
channels in Palca, Bolivia, which is described in box 3.3 in chapter 3.
Source: Kronik and Verner 2010.
as a “situation that exists when all people, at all times, have physical,
social, and economic access to sufficient, safe, and nutritious food that
meets their dietary needs and food preferences for an active and healthy
life” (FAO 2002). Though the proportion of undernourished people in
the region has generally been decreasing over the past two decades,
hunger and malnutrition remain a major problem among the rural poor,
particularly among those with little or no access to land.3
Projections of the biophysical effects of climate change and variability
on hunger in the LAC region suggest that an additional 1 million people
could be short of food by 2020 (Parry and others 2004). And in Brazil, sim-
ulations indicate that an average reduction of 18 percent in agricultural
output could lead to an increase in rural poverty of 2 to 3.2 percentage
points (De la Torre, Fajnzylber, and Nash 2009).
Climate change may also affect food security indirectly, notably by
increasing the demand for bioenergy in an effort to reduce reliance on
fossil fuels, affecting world food prices. Increasing the use of energy
from biological sources presents both opportunities and risks for food
security. It could create new market opportunities for agriculture, fos-
ter rural development, and alleviate poverty, not least by improving
rural access to sustainable energy. But if not managed sustainably, it
could seriously threaten food security, especially for some of the most
vulnerable people. Rising demand may raise food prices and increase
food price volatility, compounding the food-price effects of more fre-
quent floods and droughts across the world. Higher food prices will
have both positive and negative effects on rural poverty, affecting food
producers and food consumers in opposite directions. The net effect
in most low-income countries, however, including those in Latin
America and the Caribbean, is to increase the poverty rate (Ivanic and
Martin 2008).
It must be said that the socioeconomic environment is a more important
determinant of food security than are the biophysical impacts of climate
change (Tubiello and Fischer 2007). Much depends on the socioeco-
nomic status of a country or a population group prior to climate change
(FAO 2006). In a poor agrarian society, a failure in agricultural produc-
tion exacerbates poverty, malnutrition, and hunger for all, not just
those whose primary activity is agriculture. It also increases depend-
ency on other sources of income and livelihood. Recurrent production
failures erode livelihood assets, exacerbating poverty and eventually
leading to migration. Much will therefore depend on implementation
of pro-poor policies to combat these effects.
110 Olesen
assets differently, but when several assets are simultaneously damaged, for
instance, by a hurricane, it may be insufficient. It also may not be an
option for people who lack sufficient land or a relatively stable social
structure or who have insufficient human capital.
Communal pooling distributes risks across households, perhaps
through joint ownership of assets or through sharing labor or incomes
across households. It may also involve mobilization of resources that are
held collectively during times of scarcity. A kind of communal pooling
that draws heavily on social capital is the formation of cooperatives, to
improve farmers’ access to markets and improve the provision of inputs
to offset a less-favorable climate, for example, by organizing water har-
vesting and irrigation facilities.
Exchange can substitute for the first four of these strategies if the rural
poor have access to markets. The sale of farm products for cash allows
people to purchase other assets they need or to save for a time of greater
need. For people not fortunate enough to hold such savings, a crisis such
as a drought may force them to resort to exchange of assets such as live-
stock or household items for basic necessities to survive. To be success-
ful, however, exchange often requires considerable specialization and
institutionalization of exchange relations. Buying insurance to cover crop
or property losses from droughts or floods would fall into this category
of specialized and institutionalized exchange. The use of insurance against
crop failure varies considerably among LAC countries. On average across
countries, 6 percent of cropland is covered by insurance. The figure in
Uruguay is 20 percent (Edmeades and others 2008). In Mexico, the use
of insurance declined considerably when the agricultural insurance mar-
ket was privatized, and smallholder farmers have had difficulty paying
insurance premiums. A more extreme climate with heavy storms, in com-
bination with mismanagement of slopes and rivers, will threaten infra-
structure that is critical for ensuring market access. Rural development
therefore needs to give priority to infrastructure that can withstand more
intense rainstorms.
These strategies typically draw on several of a household’s available
assets simultaneously. Agrawal and Perrin (2008) note that the most com-
mon adaptive strategies are diversification and communal pooling, and a
combination of diversification plus exchange.
Box 4.5
adding diversity to the crop rotation and improving soil and water
resources.
Changes in farming practices, like the broader changes in family and
community livelihood strategies that were outlined in the preceding sec-
tion, will need to draw on and strengthen the assets available. Changing
the timing and location of agricultural activities draws on social capital,
using available labor to make better use of rainfall variations. That often
means intensive periods of soil cultivation, planting, and harvesting,
114 Olesen
Box 4.6
Research Perspectives
The current research focus on studying the effects of extreme events on
poor rural people in developing countries needs to move toward studying
how climate change and variability affect the long-term sustainability of
agricultural systems in marginal environments. Increased attention should
also be given to how complex agrarian farming systems can adapt to
changes in climate and climatic variability. Research should also address
the coping capacity of different regions and different social groups, par-
ticularly with respect to how migration and livelihood diversification
(and support for diversification) may assist adaptation to climate change.
Such research will not only improve knowledge of impacts, but most
importantly, it will help in building adaptive capacity at all levels within
the farming community.
118 Olesen
Notes
1. The ejido is a traditional system of communal landholding that combines
collectively used areas with individually assigned plots of land.
2. The A2 emission scenario envisages a mean temperature change of 3.4°C by
2090–99 relative to 1980–99; see IPCC 2001.
3. For many of the extremely poor, improving nutrition is the first priority if
more resources become available in the household (World Bank 2007b).
References
Agrawal, A. 2008. “The Role of Local Institutions in Adaptation to Climate
Change.” Paper presented at Workshop on Social Dimensions of Climate
Change, Social Development Department, World Bank, Washington, DC.
Agrawal, A., and N. Perrin. 2008. Climate Adaptation, Local Institutions, and Rural
Livelihoods. IFPRI Working Paper W08I-6. Washington, DC: International
Food Policy Research Institute.
Bai, Z. G., D. L. Dent, L. Olsson, and M. E. Schaepman. 2008. “Proxy Land
Assessment of Land Degradation.” Soil Use and Management 24: 223–24.
Berdegue, J. A., and H. M. Ravnborg. 2007. Agricultural Development for Poverty
Reduction: Some Options in Support of Public Policy Interventions. Copenhagen:
Danish Institute for International Studies.
Binswanger, H. P. 2007. “Empowering Rural People for Their Own Development.”
Agricultural Economics 37: 13–27.
Agrarian Livelihoods and Climate Change 119
123
124 Nielsen
impossible. Also, in some cases corals that were presumed lost, following
bleaching events, have revitalized after some years, making it difficult to
pronounce with certainty the extent of permanent loss. One meta-analysis
of 65 studies covering 263 sites in the Caribbean found that between
1977 and 2001 the average hard coral cover on reefs was reduced by
80 percent, from about 50 percent to 10 percent. Table 5.1 shows the sta-
tus of coral reefs in Latin America and the Caribbean as reported by the
Global Coral Reef Monitoring Network in 2008. According to that report,
around 36 percent of coral reefs are effectively lost or are at a critical stage,
and 24 percent are threatened. Only 39 percent appear not to be under
significant threat. These findings are consistent with those of Burke and
Maidens (2004), who identified more than 80 percent of the reefs in
Jamaica, Haiti, and the Dominican Republic as threatened by human activ-
ities, with one-third under very high threat. However, these data are associ-
ated with great uncertainty and specifically do not consider the future
threat posed by climate change. Recent World Bank studies using the
Intergovernmental Panel on Climate Change (IPCC) business-as-usual sce-
nario (and a low-temperature sensitivity scenario) predict that coral reefs in
the Caribbean will disappear completely between 2060 and 2070 under
these external impacts, assuming no adaptation measures are implemented.
More than half of the region’s inventory of mangroves is classified as
endangered. Mangrove loss is primarily due to clearing and changes
in hydrology associated with reclamation for aquaculture and coastal
development. Mangroves are threatened by sea level rise in areas where the
coastline is built up, making their landward retreat impossible, and by
Table 5.1 Status of Coral Reefs in Latin America and the Caribbean
Reefs Reefs at Reefs at
Coral reef Effectively at critical threatened low threat
Region area (km2) lost reefs (%)a stage (%)b stagec level (%)d
North Caribbean 9,800 12 13 30 45
Central America 4,630 14 24 22 40
Lesser Antilles 1,920 13 31 22 34
S Tropical America 5,120 13 40 17 30
Total 21,470 13 23 24 39
Source: Wilkinson 2008.
Note: Coral reef area from the World Atlas of Coral Reefs (2001). Estimates determined using considerable coral reef
monitoring data, anecdotal reports, and the expert opinion of hundreds of people associated with the Global
Coral Reef Monitoring Network. Assessments should be regarded as indicative because of insufficient data for
many regions.
a. Reefs “effectively lost” are 90 percent lost and unlikely to recover soon.
b. Reefs at “critical stage” have 50 percent to 90 percent loss of corals.
c. “Threatened” reefs, are reefs showing moderate signs of damage: 20 percent to 50 percent loss of corals.
d. “Low threat”–level reefs are under no immediate threat of significant losses (except from global climate change).
Coastal Livelihoods and Climate Change 127
Box 5.1
(continued)
128 Nielsen
Acidification occurs once availability of carbonate ions in the ocean starts to de-
crease because more CO2 is dissolved in the ocean. Carbonate ions are vital
skeletal building blocks for reef-building corals. Once carbonate levels fall below
a certain threshold, corals lose their ability to calcify, eroding coral reef structures
and the multitude of services they provide (Hoeg-Guldberg and others 2007;
McField and Kramer 2007; Wilkinson and Souter 2008).
In 1998 and again in 2005, increases in sea surface temperature in the
Caribbean caused severe bleaching of corals, weakening and killing many. For
mangroves, the possible change in the saltwater/freshwater balance will alter
productivity, and temperature changes are expected to decrease photosynthe-
sis and impair root structures and seedlings. Already, large storms have elimi-
nated 10 mangrove forests in the past 50 years in the Caribbean (McLeod and
Salm 2006), and in Antigua and Barbuda a 3–4-millimeter rise in sea level has
caused a 2 percent annual rate of mangrove loss. If sea level rise continues at
this rate, other things equal the island would have no mangroves left by 2075
(Murray n.d.)
Millions of people in communities that rely on mangroves and coral reefs for
fisheries and tourism will be affected by their deterioration, which will wipe out
the amenities that these ecosystems provide, such as absorbing carbon dioxide
and acting as filtering systems for nutrients. The deterioration will decrease
coastal protection from storms and floods, diminish the outstanding biodiversity
that people come to view, and decrease the availability of the fish that tourists—
and locals—consume. Tourist communities throughout the Caribbean and in the
Pacific count on coral reefs to attract snorkelers, scuba divers, and other tourists.
In addition, small communities of poor artisanal fishers will feel the impacts
because of their reliance on near-shore, demersal fish species, whose numbers
may decline as climate change takes its toll on the reefs and mangroves on which
these species rely on for spawning and hiding grounds.
According to the World Rainforest Movement (WRM 2002), annual productiv-
ity per hectare of mangroves is estimated at 1,100 to 111,800 kilograms of fishery
catches, and it is 10 to 370 kilograms per hectare for corals. The annual value of
mangroves varies from US$900 to $12,400 per hectare. These numbers only take
into account fisheries and none of the other services that mangroves provide,
including recreation and tourism. If they are included in calculations, products
and services of mangroves are valued from US$ 200,000 to $900,000 per hectare
annually (McLeod and Salm 2006). The accompanying table shows the potential
loss to coastal industries from a major loss of coral.
A recent World Bank study found that around 65 percent of all species in the
Caribbean rely to some extent on coral reefs for survival. Thus if corals collapse, so
may the species and overall ecology in the area that depends on them (De la
Torre, Fajnzylber, and Nash 2009). Studies have shown that some species of man-
groves and coral reefs have reactive adaptation capacities; thus speculations exist
whether adaptation over time is possible. However, due to the severe pressure
from human activities, most scientists agree that prolonged climate trends will
impair mangrove and coral productivity to a point of no recovery unless serious
measures are taken to protect them.
Source: Author.
Mangroves and coral reefs play vital roles in supporting the livelihoods
of millions of people who depend on them for fisheries and tourism, and
they help to weaken the force of storm surges. Coral reefs and mangroves
are most abundant throughout the Caribbean Sea, the Pacific Central
American coast, and the northern coast of South America where they act
as nurseries to a large range of marine species. They are vital to the attrac-
tiveness of these coasts as tourist destinations and hold great potential for
pharmaceutical uses. For centuries, coastal populations in Latin America
and the Caribbean have relied on them for survival. Any decline—or fur-
ther decline—caused by climate change will significantly affect people’s
livelihoods.
130 Nielsen
Box 5.2
what may be lost in fisheries. Moreover, with the expected impacts of cli-
mate change on farming, many communities may be hit from two sides
(see chapter 4).
Fishermen’s vulnerability to climate change depends greatly on the
quality of national and international governance of available assets such as
access to insurance and land and sea rights. Small coastal and artisanal
fisheries in Latin America and the Caribbean have long faced challenges
related to the globalization of trade, privatization of access rights to fish-
eries, and poorly functioning natural resources protection systems. They
also often lack democratic representation, which prevents small-scale
fishery organizations from participating effectively in national decision
making (Allison and others 2005).
First, a lack of property rights to fisheries causes particular difficulty.
Though some communities have relied on an area for fishing for centuries,
the common-property nature of most fisheries leaves the area open to any-
one. For example, property rights to protect mangroves from conversion to
shrimp farms and other aquaculture are rare, and so small-scale fishermen
have often been overrun by large-scale industrial aquaculture. Aquaculture
is a fast-growing industry in Latin America and the Caribbean, but poor
local fishermen have benefited little from it and have often lost their rights
to fish in certain areas because of it. In Placencia, Belize, for example, local
communities are fighting to protect their mangroves from the intrusion of
large-scale shrimp farmers. The same is occurring in Cispata Bay, Colombia,
where pollution from aquaculture has severely affected water quality in
nearby agricultural zones that support roughly 2,500 families, none of
whom obtains any financial benefits from the farms.
Second, small-scale fishermen often lack access to sufficient insur-
ance. Particularly, they are unable to obtain any insurance that protects
against losses following a disaster. Insurance is too expensive because of
the high risk associated with their work and the poor safety of their boats
(Allison and others 2005; McGoodwin 2001).
Finally, because many small-scale fishing communities are remote from
urban areas, early warning systems have been lacking. Fishing communities,
particularly those of indigenous origin, receive little early warning before a
disaster strikes. In 2007, Hurricane Felix struck Nicaragua and Honduras,
wiping away farmlands, destroying vital fishing grounds, and affecting the
lives of about 198,000 people. As described in chapter 4, those affected
were primarily from Miskito communities that had received little warning
before the hurricane hit.
Overall, the fishing communities of Latin America and the Caribbean
are seriously threatened by climate change. It is important to note, however,
136 Nielsen
Tourism
Scholars and researchers agree that the tourism industry in Latin America
and the Caribbean will be damaged by climate change and variability if
no adaptation measures are implemented. Overall, researchers expect a
northward shift in the tourist industry away from the region (Ehmer and
Heymann 2008; UNWTO 2008; see figure 5.1). Increased threats from
natural disasters, torrential rainfall, and elevated temperatures—possibly
coupled with higher travel costs reflecting government policies to miti-
gate global warming—may prompt travelers to vacation closer to home.13
Unfortunately, little research has been conducted on the causal relation-
ship between climate change and tourism or on how climate change
might affect people working in the industry. Even so, it is clear that
because of the climate-sensitive nature of tourism and the increasing
reliance on it in the LAC region, safeguarding tourism against climate
change is strongly called for.
Overall, roughly 40.7 million people in the LAC region are employed
in primary tourist-related activities, and 23.5 million in secondary ones.14
Though people working in tourism usually have higher standards of living
and more education than those in fishing and farming communities, the
industry employs many illegal immigrants and is so closely connected to
other industries that employ the poor that it deserves much attention in
the debate on climate change and livelihoods (Aguayo, Exposito, and
Nelida 2001; ECLAC 2003b). For example, in the Galapagos thousands
of illegal migrant workers from mainland Ecuador have been arriving for
decades to work as maids, waiters, cleaners, and shop assistants (Carrol
2008). Poor people who work in any industry related to tourism, or
impoverished communities that wish to develop tourism, may be affected
by climate change and variability. Tourism is one of the main develop-
ment tools used to raise the economic status of small communities.
Indeed it has been a successful tool, but it has been developed without
climate change considerations, making it vulnerable to climate change.
In recent decades, the tourist industry has grown significantly in nearly
all LAC countries in Latin America and the Caribbean, in terms of both
contribution to GDP and employment.15 Though the Caribbean and
Mexico hold the highest market shares (14.4 percent and 8.2 percent,
respectively), poorer countries with small market shares have experienced
Figure 5.1 Impacts of Climate Change and Variability on the Tourist Sector in Latin America and the Caribbean
IBRD 37784
MAY 2010
the greatest growth in both employment and value (table 5.4). For exam-
ple, as table 5.4 shows, Nicaragua, with only 0.5 percent of the market, saw
an annual growth in tourism of nearly 21 percent from 1990 to 2005, rais-
ing the value of the industry from US$12 million to US$207 million.
Employment has grown at an annual rate of 8.4 percent in Nicaragua’s
primary tourist-related activities and 0.9 percent in secondary tourist-related
industries (table 5.5). The trend is quite clear; poor countries in particular
are benefiting from increased tourism. Unless climate safeguard measures
are implemented, these benefits may come under threat.
The tourist industry of Latin America and the Caribbean will feel the
impacts of climate change and variability through three pathways.
Table 5.4 Value of the Tourist Industry in Selected Countries in the Americas,
1990–2005
Average annual Market
US$millions growth (%) share (%)a
Country 1990 2000 2005 1990–2005 2004
Belize 44 116 b 8.2 0.2
133
Bolivia 55 68 205 9.2 0.3
Brazil 1,492 1,810 3,861 6.5 3.8
Caribbean 8,721 17,145 20,386 5.8 14.4
Colombia 406 1,030 1,218 7.6 0.6
Mexico 5,526 8,294 11,803 5.2 8.2
Nicaragua 12 129 207 20.9 0.5
Peru 217 837 1,308 12.7 1.0
Source: UNWTO 2008.
Note: Includes values based on directly related tourist activities, such as hotels, restaurants, entertainment, and
tours.
a. The United States holds almost 56 percent of the market share.
b. Most recent reported numbers are from 2004.
Coastal Livelihoods and Climate Change 139
restaurants, bars, and gift shops. In the six weeks that followed the event,
revenue loss amounted to US$7 million (Jackson n.d.). In the Bahamas—
one of the Caribbean’s most popular tourist destinations—most shops at
the international market in Freeport were still closed in mid-2009, follow-
ing the landfall of Hurricane Hanna in September 2008.16 In Grenada,
Hurricane Ivan in 2004 caused damage to more than 90 percent of all
hotels, totaling US$128 million, or 29 percent of GDP. The hurricane
140 Nielsen
severely hurt ecotourism and cultural heritage sites, accounting for more
than 60 percent of job losses in this subsector (Becken and Hay 2007).
Box 5.3
Box 5.4
Merida
Cancun
IBRD 37785
MAY 2010
Indeed, storms have historically posed the most significant threat to the area.
In 1988, Hurricane Gilbert swept away about half of the beaches in Cancun, and
in 2005 Hurricane Wilma took roughly 8 miles of sandy beach shore. Wilma also
significantly discouraged tourism for the 2006 season and severely damaged
(continued)
Coastal Livelihoods and Climate Change 143
hotels, which were forced to stay closed for months. The total cost of Wilma was
US$17.2 million, and overall more than 16,000 rooms, or 60 percent of capacity,
were damaged (Hendershot 2006; Martinez 2008). That year the number of jobs
in primary tourist activities fell by 234,200 (ILO 2008). As climate change and vari-
ability intensify, the Yucatan peninsula is almost certain to feel similar and more
long-term effects.
Source: Author compilation.
Figure 5.2 Effects of Climate Change on the Economy, the Tourist Industry,
and Livelihoods
Erosion of
physical and
natural assets Erosion of
through loss of financial assets
Climate change infrastructure and due to decreases
and variability destruction of the in money spent
environment in the area and
the cost of
rebuilding; Businesses suffer
further loss in losses and may
Loss of visitors visitor numbers. close.
Source: Author.
Nevis, and Turks and Caicos, all nations where income from tourism
makes up a large share of GDP. Cuba could face an economic impact of
13 percent loss of GDP by midcentury and a 27 percent loss by 2100
(Bueno and others 2008). It is clear that such losses in the tourist indus-
try would significantly affect both the economy and society as a whole
in the individual countries.
In many cases, the potential for damage from climate change is greater
because of past disregard for environmental sustainability in the develop-
ment of the tourist industry. The dense concentration of hotels in coastal
areas has been pushed by national development policies responding to
the preference of developers and visitors for beachside properties
(Jackson n.d.). In addition, because beachside tourism has remained a
prime choice for visitors, beaches have been poorly protected from devel-
opment and abuse. Developers have been allowed to create white, sandy
beaches even where these are not natural and may cause damage. For
example, converting the unique lava stone beaches in St. Vincent and the
Grenadines to white, sandy ones may worsen the country’s existing ero-
sion problems.18 Building beachside cabanas in the path of hurricanes, as
well as in other low-lying coastal areas, may cause significant safety issues
and long-term economic problems in the event of storms, flooding, and
rising sea level.
Coastal Livelihoods and Climate Change 145
Box 5.5
Janeiro have no sewerage connection. Even when such services are avail-
able, many of the urban poor find their cost prohibitive. For example, in
Guatemala, where services are in principle available to the poor in most
cities, 20 percent to 40 percent of the poor do not connect to them
because they find the cost too high.
Few of the urban poor have the opportunity to attend school beyond
the primary level, often because the distance they need to travel to
reach secondary school is too long. Their lack of sufficient education
confines them to work in the informal sector, which offers little or no
social insurance or security. Only about 20 percent of the urban poor
work in the formal sector regionwide. Only 6 percent of the urban poor
in Mexico, and as few as 4 percent in Peru have pension entitlements
(Fay 2005).
All in all, the urban poor are very vulnerable both to climate-related
disasters and to the effects of ongoing climate change. Intensifying hur-
ricanes, torrential rainfall, landslides, and other climate-related events
may cause severe damage to houses or destroy them altogether. Such
events also destroy vital roads and utilities, often cutting off poor
neighborhoods from help. A lack of water and sanitation infrastructure
encourages disease outbreaks, especially outbreaks of diarrheal dis-
eases. Warmer mean temperatures and standing water following hurri-
canes and floods also raise the prevalence of diseases such as Chagas
disease, dengue fever, and malaria in urban areas—which is worrisome,
considering the cramped conditions and overall lack of health care in the
urban slums. Further, it can be assumed that disease outbreaks and the
rebuilding of houses and other infrastructure will keep individuals away
from work, perhaps depleting their savings and credit (financial capital).
And when adults fall ill, younger family members may enter the work-
force, compromising their education and later earning prospects
(decreasing their human capital).
Lack of social and cultural capital is likely to worsen the situation for
urban slum dwellers. Social networks, familiar ties, and traditions are
much more integrated in rural areas, where individuals share the same
culture and personal attachment to the community (Verner and Egset
2008). Once migrants arrive in cities, they tend to lose their community
and cultural identity and find it difficult to establish a personal network
that offers the same level of security. Although some social capital exists
in urban areas, Fay (2005) has established that social ties are more indi-
vidualistic and based less on familial bonding. In addition, the networks
that are established tend to be unstable because of the mobility of people
148 Nielsen
in urban areas, high crime rates, and corruption. The lack of social capital
may have very negative effects after natural disasters, especially consider-
ing the existing high crime rates in megacities.21
To sum up, the urban poor are highly vulnerable to climate change
and variability because they lack the assets to cope with them,
because of the vulnerable lands on which they live, and because of the
density of their living quarters. The lack of such assets as clean water,
sanitation, electricity, social services, safe land, insurance, and secure
employment significantly limits their human, physical, and financial
capital, and they lack some of the important social and cultural capital
often found in rural areas. During a natural catastrophe, the lack of
these assets is likely to worsen their situation, and capital may continue
to deteriorate. It is clear that to help the urban poor cope with either
short-term disasters or long-term climate change affecting health and
work conditions, building their livelihood assets is vital. That implies
the governance structures affecting the distribution of assets need to
change.
Box 5.6
(continued)
152 Nielsen
Rangers in PHMR.
Photo: Sara Trab Nielsen.
connection to the area, and (c) as fish resources became scarce, alternative
employment for fishermen was needed.
In the past eight years, TIDE has led a successful initiative helping fishermen to
diversify their livelihoods. Fishermen are trained as rangers and marine specialists,
who help monitor fish stocks and diversity in the reserve. In addition, TIDE Tours
was established to promote ecotourism in the region. It offers canoe trips up the
Rio Grande, boat trips through the mangroves, fly-fishing, catch and release, and
snorkeling tours, all guided by local people who previously worked as fishermen.
According to testimonials from several people employed by TIDE, their new jobs
as tour guides, taxi drivers, captains, and rangers provide a much more sustain-
able and reliable livelihood than fishing.
Overall, the reserve is a success. Though TIDE occasionally catches fishermen
in the reserve during the off season, the institute generally receives positive sup-
port from the community, and species diversity as well as fish stocks have
increased in recent years. Even manatees can now be spotted in the reserve.
Source: Author.
attractions such as sea turtles and coral reefs may disappear. Such changes
require not only development of alternative tourist attractions but also a
focus on new marketing strategies. Enhancing cultural tourism may be an
alternative. All countries and regions within them have their own unique
cultural traditions and traits that may be used to attract visitors.
in safer areas and should inform residents about the importance of their
vacating high-risk areas. Relocation programs have not always suc-
ceeded, because they are costly to the people affected, especially if
they entail moving away from their means of making a living.29 In St.
Vincent, for example, the government has found that individuals do
not accept relocation packages because they are reluctant to leave a
house that essentially costs them nothing—even if it lies in a high-risk
area. Some families who agree to be relocated leave some of their
members behind in the risk-prone house, in fear that someone else
may take it.30 Indeed, to prevent people from remaining in high-risk
settlements, or discourage new people from moving in, the areas
should be converted to alternative uses precluding habitation, such as
urban agriculture, parks, or sports and playing fields. Informing people
about the risk associated with staying is crucial to success, as is making
sure that the substitute housing is affordable and offers some of the
amenities that are lacking in the high-risk location, such as clean water,
sanitation, and electricity. Relocating people to neighborhoods with serv-
ices may indeed cost less in the long term, while protecting people
against losing their livelihoods.
Finally, it is vital to improve services that enhance human capital, such
as education, health care, and other social services. Providing better edu-
cation benefits society as a whole. More educated people can help
improve the overall standard of living, provided enough jobs are available.
Jobs should come with social benefits such as better health services.
People in better health are more resilient to disease outbreaks, and pub-
lic health measures during a disease outbreak help to prevent deaths
(chapter 6).
Research Perspectives
Overall, the impacts of climate change phenomena on coastal liveli-
hoods remain severely understudied. More detailed information is
needed on each subregion, country, locality, and sector. In particular,
up-to-date time-series data on employment and the value of coastal
sectors, as well as an overview of climate-related shocks and trends in
the different countries, are vital. International organizations, non-
governmental organizations, and individual governments should con-
tinue to fund research and take initiatives to improve the livelihoods
of impoverished coastal populations to safeguard them against climate
change and variability.
Coastal Livelihoods and Climate Change 157
Notes
1. The complete LAC region has a population of 556 million (Hinrichson 2008;
World Bank 2008).
2. Low-elevation coastal zones (LECZ) are the areas along the coast that are less
than 10 meters above sea level and will be most hit by climate change events
and variability (McGranahan, Balk, and Anderson 2007).
3. The livelihood assets comprise physical, financial, human, social, cultural, and
natural capital.
4. Only a little information will be drawn from the agricultural sector since it is
covered in chapter 4. The science behind climate change is discussed in
appendix A.
5. One of the main issues is that large fish populations and predator species have
been replaced in the catch by smaller, less valuable species, or juveniles of the
species traditionally caught. Thus it can be assumed that while the volume of
158 Nielsen
fish caught may increase for a time, fish stocks are being serially depleted as
effort transfers from one species to another, down the food chain.
6. Fishers also noted that when gasoline prices began to rise, several of them
were forced to stay on land, as they were no longer able to obtain a catch suf-
ficient to cover the cost of their fuel. Information obtained through interviews
with Dennis Garbutt from EarthWatch Belize and local fishers in Monkey
River, Belize, 2008.
7. Tourism is included in this analysis because of its climate-sensitive nature, the
increase in tourism development in smaller communities, and in particular the
increase in development of ecotourism.
8. Number based on estimations compiled from FAO databases. The number
includes artisanal, inland fisheries, marine fisheries and all people employ-
ment in secondary industries such as marketing and processing. However, it
excludes the vast number of unregistered fishermen. It is difficult to obtain
accurate figures for how many people work in fisheries in the LAC region and
worldwide, as millions are unregistered (see Allison and others 2009, 9).
9. Honduras is an exception, as it has a well-developed industrial shrimp fleet
and is a major exporter of shrimp.
10. The study conducted by Allison and others (2009) included 132 countries
worldwide. Countries were rated based on their possible climate change
exposure, their economies’ dependence on fisheries, and their ability to adapt
to climate change.
11. With the exception of shrimp, the large-scale industrial fleet does not fish
shallow-water demersal species, such as those found adjacent to coral reefs.
Rather they fish a few pelagic species far off shore. It is important to note
that climate change and variability will affect demersal and pelagic species
differently.
12. Information obtained from fieldwork in Belize in 2008.
13. Climate change effects such as rising temperatures may simply result in
changing tourist seasons overall, for example, with the high season for travel
beginning and ending later in the year. Not enough local climate data exist
to be able to predict these possible changes and their impact (Ehmer and
Heymann 2008).
14. This discussion splits the tourist industry into two categories: (a) primary
tourist-related jobs, which are directly related to tourist activities, including
jobs in hotels, restaurants, and wholesale and retail trade, and (b) secondary
tourist-related jobs, including anything that may be affected by turmoil in the
primary industry, such as construction, transport, communications and non-
classifiable jobs, usually in the informal sector.
15. The rapidly expanding tourist industry in the LAC region can be separated
into three categories: (a) mass tourism such as cruise liners, (b) large scale “sun
Coastal Livelihoods and Climate Change 159
and surf” tourism, and (c) sustainable ecotourism, including cultural tourism.
“Sun and surf” tourism refers to the classic sort of tourism consisting of long
days at the beach, shopping, and nights out, with little regard for the environ-
ment; ecotourism consists of visitors who travel to see the local environment
and learn about environmental protection. However, the two different indus-
tries are starting to overlap, as ecotourism now also includes beach travel in
which tourists can stay in sustainably built and kept resorts, and regular
hotels are starting to incorporate environmental measures. “Mass tourism”
refers mainly to cruise liners that travel from port to port. Because cruise lin-
ers provide travelers with all their basic needs, hotels, restaurants, stores, and
local guides and communities profit little from the industry, and so mass
tourism is not considered in this analysis.
16. Information obtained during fieldtrip to The Bahamas, May 2009.
17. Tourism, more than any other industry, faces indirect impacts from climate
change mitigation measures. For example, climate mitigation bid the arrival of
carbon taxation systems, which may raise fuel prices and the price of plane
tickets. Currently, most airlines give travelers a choice of whether to purchase
carbon offsets or not. However, it is to be expected that a global carbon tax
will be implemented on transport costs in the near future. That, coupled with
rising fuel prices, may drive travel costs to a level where people simply opt out
of traveling. That, according to the CHA-CTO (2007) could jeopardize the
sustainable livelihoods of about one-third of the Caribbean population.
18. Currently many of these beaches are being destroyed by removal of stones for
construction purposes. Most of this industry is informal. Observations from
field trip to St. Vincent, 2008.
19. For the purpose of this chapter, “urban poverty” will refer only to megacities
and associated slums. However, the general definition of the term “urban”
includes everything from small towns to megacities (Fay 2005).
20. In Brazil these cities include Baixada Santista, Belém, Fortaleza, Grande
Vitória, Maceió, Natal, Porto Alegre, Recife, and Rio de Janeiro. Other highly
vulnerable cities include Barranquilla, Colombia; Lima, Peru; Montevideo,
Uruguay; Maracaibo, República Bolivariana de Venezuela; Belize City, Belize;
Panama City, Panama; Havana, Cuba; Santo Domingo, Dominican Republic;
and Port-au-Prince, Haiti. These cities’ vulnerability has been rated based on
level of infrastructure, economy, and population size (OECD 2007).
21. That was apparent in the United States when looting and crime soared dur-
ing the days following Hurricane Katrina.
22. Cultural identity is a vital consideration in initiatives that aim to build new
livelihoods. It should be mentioned that some cultures would not accept
aquaculture as an alternative. The choice is highly differentiated. For example,
in Belize fishing villagers were willing to develop sustainable small-scale aqua-
culture if land rights were allocated to them, while in other parts of the
160 Nielsen
References
Aguayo, Eva, Pilar Exposito, and Lawelas Nelida. 2001. “Econometric Model of
Service Sector Development and Impact of Tourism in Latin American
Countries.” Applied Econometrics and International Development 1, no. 2.
Allison, Edward H., W. Neil Adger, Marie-Caroline Badjeck, Katrina Brown,
Decian Conway, Nick K. Dulvy, Ashley Halls, Allison Perry, and John D.
Reynolds. 2005. Effects of Climate Change on the Sustainability of Capture and
Coastal Livelihoods and Climate Change 161
Ehmer, Philippe, and Eric Heymann. 2008. Climate Change and Tourism:
Where Will the Journey Lead? Deutsche Bank Research, Frankfurt am
Main.
Fay, Marianna, ed. 2005. The Urban Poor in Latin America. A World Bank Report.
Washington, DC: World Bank.
Freeman, Gary E., and Craig Fischenich. 2000. “Gabions for Streambank Erosion
Control.” U.S. Army Corps of Engineers, Environmental Laboratory. http://el
.erdc.usace.army.mil/elpubs/pdf/sr22.pdf.
FAO (Food and Agriculture Organization). 2007. Fisheries and Aquaculture
Information and Statistics Service.
———. 2008. Fisheries and Aquaculture Country Profiles. http://www.fao.org/
fishery/countryprofiles/search/en.
Hall, Michael C., and James Higham, eds. 2005. Tourism, Recreation, and Climate
Change. Channel View Publications.
Hamilton, Jacqueline M., D. J. Maddison., and R. S. J. Tol. 2005. “Climate Change
and International Tourism: A Simulation Study.” Global Environmental Change
15: 253–66.
Hendershot, Rick. 2006. “Cancun Beach Being Restored at Record Pace.” Health
Guidance for Better Health. http://www.healthguidance.org.
Hinrichson, Don. 2008. “Ocean Planet in Decline.” People and Planet: People and
Coasts and Oceans. http://www.peopleandplanet.net/doc.php?id=429&
section=6.
Hoeg-Guldberg, O., et al. 2007. “Coral Reefs under Rapid Climate Change and
Ocean Acidification.” Science 318 (December 14).
ILO (International Labor Organization). 2008. LABORSTA—Database of Labor
Statistics. http://laborsta.ilo.org.
IUCN. 2007. “Latin American Park Congress: 2008–2018 to Be ‘Decade of MPAs.’”
http://www.iucn.org/.
Jackson, Ivor. n.d. “Potential Impacts of Climate Change on Tourism.” Issues paper
prepared for OAS, Mainstreaming Adaptation to Climate Change (MACC)
Project.
Kawasaki, T. 2001. “Global Warming Could Have a Tremendous Effect on
World Fisheries Production.” In Microbehavior and Macroresults:
Proceedings of the Tenth Biennial Conference of the International Institute of
Fishery Economics and Trade, ed. R. S. Johnston and A. L. Shriver. July
10–15, 2000, Oregon, U.S.
Kelly, Kathryn, and Ronald Sanabria. 2008. “Mainstreaming Biodiversity
Conservation into Tourism through the Development and Dissemination of
Best Practices.” World Bank presentation, September 18, World Bank,
Washington, DC.
Coastal Livelihoods and Climate Change 163
Sumaila, Ussif Rashid, Sylvie Guenetta, Jackie Alder, David Pollard, and Ratana
Chuenpagdee. 1999. Marine Protected Areas and Managing Fished Ecosystems.
CMI Reports. Bergen: Chr. Michelsens Institute.
Tropical Re-Leaf Foundation. 2008. “Nariva Swamp Restoration Initiative.”
http://www.ema.co.tt/main.htm.
UNEP (United Nations Environment Program). 2007. “Climate Change Hits
Hard on Latin America and the Caribbean.” UNEP Press Release. http://www
.unep.org.
UN Habitat (United Nations Human Settlement Program). 2003. The Challenge
of Slums. Global Report on Human Settlements. London: Earthscan
Publications.
United Nations Population Division. 2006. World Urbanization Prospects: The
2005 Revision. Department of Economic and Social Affairs, Urban and Rural
Areas Dataset (POP/DB/WUP/Rev.2005/1/Table A.6).
UNWTO (United Nations World Tourism Organization). 2008. “Climate Change
and Tourism: Responding to Global Challenges.” Advance Summary.
http://www.unwto.org/index.php.
Vergara, Walter. 2005. Adapting to Climate Change: Lessons Learned, Work in
Progress, and Proposed Next Steps for the World Bank in Latin America.
Sustainable Development Working Paper 25, World Bank, Washington, DC.
Vergara, Walter, Natsuko Toba, Daniel Mira-Salawa and Alejandro Deeb. 2009.
“The Consequences of climate-induced coral loss in the Caribbean by
2050–2080.” In Assessing the Potential Consequences of Climate Destablization
in Latin America. Sustainable Development Working Paper, World Bank,
Washington DC.
Verner, Dorte, and Willy Egset. 2008. Social Resilience and State of Fragility in
Haiti. A World Bank Report. Washington, DC: World Bank.
WHO (World Health Organization). 2008. Global Epidemics and Impacts on
Cholera. http://www.who.int/topics/cholera/impact/en/index.html.
Wilkinson, C. R., ed. 2008. Status of Coral Reefs of the World: 2008. Townsville,
Australia: Global Coral Reef Monitoring Network and Rainforest Research
Center.
Wilkinson, Clive, and David Souter, eds. 2008. Status of the Caribbean Coral Reefs
After Bleaching and Hurricanes in 2005. Townsville, Australia: Global Coral
Reef Monitoring Network.
World Atlas of Coral Reefs. 2001, rev. 2007. UNEP/WCM. http://www
.unep-wcmc.org/marine/coralatlas/index.htm.
World Bank. 2001. “Attacking Brazil’s Poverty: A Poverty Report with a Focus on
Urban Poverty Reduction Policies.” Report No. 20475-BR, World Bank,
Washington, DC.
Coastal Livelihoods and Climate Change 165
Climate change and variability pose a serious health risk to many develop-
ing countries, most of which are poorly prepared to face it. By 2000, accord-
ing to estimates from the World Health Organization (WHO), climate
change and variability had caused the loss of more than 150,000 lives
around the world and 5.5 million disability-adjusted life years (DALY).1 For
Latin America and the Caribbean, the human cost of climate change and
variability by 2000 was an estimated 92,000 DALY, placing the region
among those with the most negative health effects.2 These data show that
climate change and variability have already had a significant impact on
health, as the number of cases of illness and mortality from disease and
extreme weather events has increased.
Malaria has increased in several countries in Latin America and the
Caribbean (LAC) over the last two decades, particularly in Bolivia, Brazil,
Colombia, Guyana, Peru, and Suriname. The increase has occurred
despite improvements in public health infrastructure that brought a
decrease in child mortality in every country of the region over the same
period (table 6.1). However, while malaria is likely to extend its reach
into new areas as a result of climate change, other areas can expect a
decline in the incidence of the disease, so that much uncertainty exists
about the net effect on malaria rates in the region as a whole.
167
168 Andersen, Geary, Portner, and Verner
death from
temperature- hypothermia,
heat strokes,
related morbidity
cardiorespiratory
diseases
waterborne
cholera
diseases
threatened
malnutrition
food supply
• Populations that border regions with endemic diseases that are sensi-
tive to climate conditions
• Regions for which a correlation has been observed between epidemic
diseases and weather extremes such as El Niño episodes
• Areas where several climate impacts are projected to affect health (for
example, stress on food and water supplies and risk of coastal flooding)
• Areas at risk because of environmental and socioeconomic stresses,
whose populations have little capacity to adapt (for example, areas
with poor land use or underdeveloped health infrastructure)
Vector-Borne Diseases
Vector-borne diseases, especially malaria and dengue fever, constitute a
major health risk from climate change and variability, particularly because
these diseases are highly sensitive to even small changes in temperature
and precipitation. Notably, the number of cases of both diseases has
spiked following strong El Niño years, which are characterized by changes
in temperature and precipitation. The future impact of climate change
will depend on local conditions, as each locality presents a different
threshold for the vector. One of the major risks is that malaria will spread
into highland areas as mean temperatures rise. In Africa, malaria has
begun to occur in highland areas with cooler climates where it had not
172 Andersen, Geary, Portner, and Verner
1,400
1,200
no. of cases (1,000s)
1,000
800
600
400
200
0
1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006
malaria dengue fever
Malaria. Malaria has the greatest death toll of any vector-borne disease in
the world, annually claiming more than a million lives worldwide (WHO
2005). Even small changes in temperature and precipitation could signifi-
cantly alter the distribution of this disease. The prospects include the
spread of malaria to regions not previously plagued by it, where the pop-
ulations enjoy little immunity.
There is great uncertainty about the impact that climate change and
variability will have on malaria incidence. According to the IPCC (2007),
“there is no clear evidence that malaria has been affected by climate
change in South America,” whereas the World Health Organization
(WHO) has argued that malaria is probably the vector-borne disease
that will be most affected by climate change (WHO 2003). It is agreed,
however, that climate change and variability will likely reduce the inci-
dence of malaria in some areas and increase it in others, depending on
local factors.
In the Peruvian Amazon near Iquitos, small temperature variations of
1–2°C have been found to determine the seasonality of malaria transmis-
sion (WHO 2003). As shown in figure 6.3, an inverse relationship exists
between malaria incidence and temperature. That relationship can be
explained by the fact that temperatures in this particular area are near the
upper limits of the vector’s tolerance.
174 Andersen, Geary, Portner, and Verner
10,000 29.5
9,000
29.0
8,000
7,000 28.5
P. falciparum cases
temperature (°C)
6,000
28.0
5,000
27.5
4,000
3,000 27.0
2,000
26.5
1,000
0 26.0
J F M A M J J A S O N D
months
cases temperature
Table 6.2 Expected Changes in Malaria Transmission Rates Due to Climate Change
The Amazon Andean region Central America Southern cone
Significant reduc- Increase expected Increase expected for Increased risk along
tion expected as for cities such as Nicaragua. General southern limits of range
the climate La Paz and Quito decrease expected; due to temperature
becomes too dry due to tempera- already observed in increases.
for mosquitoes. ture increases. Honduras.
Source: IPCC 2007.
176 Andersen, Geary, Portner, and Verner
Dengue fever. As table 6.3 indicates, climate change and variability are
expected to increase the incidence of dengue fever significantly; the dis-
ease is already spreading in Mexico and central South America (Warren
and others 2006). With transmission rates likely to grow by 21 percent by
the 2050s in most parts of South America,4 dengue fever could become
one of the major health risks resulting from climate change and variabil-
ity. Although dengue fever epidemics are closely correlated with changes
in temperature and precipitation, as has been observed following El Niño
years, they are also strongly influenced by socioeconomic factors and
water storage practices. As dengue fever is primarily an urban disease,
whose risk is heightened by unplanned urbanization, poor housing, and
poor public health infrastructure, it presents a major health risk for some
of the largest cities in the LAC region.
Worldwide, 3 billion people are currently at risk from dengue fever
and tens of millions are affected, resulting in at least 653,000 DALY
worldwide (WHO 2003).5 In the last few decades, the disease has
extended its reach through the LAC region (figure 6.2), particularly
affecting urban areas (WHO 2008). The rates of both dengue and dengue
hemorrhagic fever have risen in the region during the last two decades
Table 6.3 Additional Cases of Malaria and Dengue Fever under 50- and
100-Year Future Scenarios
Additional number of cases predicted
Total cases for a 6-year period
Vector-borne disease recorded 2000–05 50-year scenario 100-year scenario
P. falciparum malaria 184,350 19,098 56,901
P. vivax malaria 274,513 16,247 48,207
Dengue fever 194,330 41,296 123,445
Total 653,193 76,641 228,553
Source: Blanco and Hernandez 2009.
Human Health and Climate Change 177
(Torres and Castro 2007). Compared with the 6-year period from 2000
until 2005, when 194,330 cases were recorded, the first quarter of 2008
alone saw more than 100,000 registered cases and 40 fatalities, most of
which occurred in Brazil (PAHO 2008).
Dengue fever is spread by mosquitoes, primarily the Aedes aegypti
mosquito, which breeds in urban environments in water-holding con-
tainers.6 The disease is seasonal and predominates during warmer, more
humid weather. Rainfall also appears to affect its transmission potential.
For the correlation between rainfall and temperature, on one hand, and
dengue epidemics on the other, a statistically significant lag has been
found of 3–7 months for temperature and 1–3 months for rainfall
(Chen 2006).
As shown in figure 6.4, the Aedes aegypti mosquito had been eradicated
in most of Latin America and the Caribbean by 1970, but by 2001 it had
made a significant comeback. The key reason was the massive, uncon-
trolled urbanization that took place in tropical America during the 1970s
and 1980s, coupled with poor maintenance of public health infrastructure.
The mosquito’s spread through much of Central and South America has
put many countries at risk of dengue epidemics. Moreover, new virus
strains have been detected in new areas. That increases the potential of sec-
ondary dengue infections, which are one factor leading to the more serious
Aedes aegypti
distribution in the
Americas: 1930s,
1970 and 2001.
IBRD 37786
MAY 2010
dengue hemorrhagic fever, which causes both internal and external bleed-
ing (Gubler 2002; pers. com. 2009).
Mexico is increasingly plagued by dengue fever—the number of
cases there has increased by 600 percent since 2001—and by dengue
hemorrhagic fever (AP 2007). Increased precipitation has been cited as
a factor. According to Joel Navarette, of the Mexican Social Security
Institute, dengue fever has changed from being a seasonal to a year-
round phenomenon in parts of the country. This corresponds with pro-
jections by the IPCC that climate change and variability will cause an
increase in the number of dengue cases.
As with malaria, migration and other population movements have
played an important role in spreading different strains of dengue fever
across Mexico. A person can develop immunity to one strain, but subse-
quent exposure to a different strain increases his or her risk of developing
the more dangerous hemorrhagic type (AP 2007).
In urban areas, change in precipitation might be a less important
causative factor than changes in water storage practices. Droughts lead
households to store water in tanks and other containers, which can serve
as breeding sites.7 As mentioned previously, dengue fever is primarily an
urban disease that is greatly influenced by socioeconomic factors such as
housing, public health infrastructure, lack of basic sanitation, and water
storage practices (Torres and Castro 2007). The importance of socioeco-
nomic factors is suggested by comparing the incidence of dengue fever on
the Texan and Mexican sides of the U.S.-Mexican border: between 1980
and 1996, 50,333 cases were reported in the three Mexican states border-
ing Texas, compared to only 43 cases in Texas (WHO 2003).
Population growth and unplanned urbanization are thought to be
important contributors to dengue fever, as is increased travel, which can
bring the disease to areas previously free from it (Torres and Castro 2007).
In the LAC region, São Paulo, Caracas, Mexico City, and Buenos Aires
are just some of the cities with unplanned growth that are at risk from
epidemics (WHO 2003).
Waterborne Diseases
Waterborne diseases present another serious health risk in Latin America,
and climate change and variability are likely to increase their incidence.
Evidence from El Niño years suggests that cholera and other waterborne
diseases are highly sensitive to changes in climate. Diarrheal diseases are
both the most common type of waterborne illness and by far the dead-
liest group of diseases associated with the environment. In Latin America
Human Health and Climate Change 179
and the Caribbean in 2002 they killed almost 20 times as many people as
malaria and dengue fever combined (Warren and others 2006). Extreme
weather events, such as floods and heavy rains, increase the incidence of
waterborne diseases, but other important causes are the risky water stor-
age and water use practices of people who lack access to clean water and
sanitation. Accordingly, the risk of waterborne diseases is expected to
increase as a result of climate change and variability—an expectation that
is supported by the evidence of El Niño’s influence on diarrheal diseases.
Evidence is ample of the link between extreme weather events and
waterborne diseases is ample. Following Hurricane Mitch, the number of
cholera cases was four times higher than normal in Guatemala and six
times higher in Nicaragua (IPCC 2001b). On the Pacific Coast of South
America, outbreaks of cholera have been tied to extreme weather caused
by El Niño; about 11,700 people died from cholera in Peru between 1991
and 1996. Similarly, when temperatures were higher than normal during
El Niño in 1997–98, an unusually large number of people suffering from
diarrhea and dehydration were admitted to hospitals for rehydration
treatment in Lima. Indeed, a time-series analysis showed an 8 percent
increase in admissions due to diarrhea for each 1°C increase in tempera-
ture (WHO 2003).
In tropical climates diarrheal diseases normally peak during the rainy
season, and studies from the United States have shown that about half of
all waterborne diseases can be attributed to extreme rainfall (WHO
2003). Most of the waterborne microorganisms that cause disease derive
from human or animal feces, which heavy rains and flooding can trans-
port into groundwater as well as surface and well water. Sources of drink-
ing water can thus become contaminated with E. coli bacteria and other
microorganisms. Communities that have combined systems for storm
water drainage and sewage run the risk of accidental releases of raw
sewage. Heavy rainfall can lead to surface runoff that allows pathogens to
enter aquifers, wells, and drinking water. Because one of the projected
aspects of climate change is more frequent heavy rains, such contamina-
tion is likely to escalate, causing a surge in the incidence of waterborne
diseases. However, underlining the complexity of forecasting the health
impact of climate change, the seasonal pattern of cholera outbreaks in the
Amazon basin has been associated with lower river flow in the dry sea-
son (Gerolomo and Penna 1999).
Rising temperatures may lengthen the season, or alter the geographical
distribution, of waterborne diseases, inasmuch as most of the protozoa,
viruses, and bacteria that cause them thrive in warm water and weather.
180 Andersen, Geary, Portner, and Verner
Malnutrition
The World Health Organization considers malnutrition the single most
important risk to global health, making up 15 percent of the total disease
burden in DALY (WHO 2003). According to the IPCC, the additional
number of people in Latin America at risk of suffering hunger in 2020 will
be 5 million under the IPCC’s A2 emission scenario (IPCC 2001a).8
However that projection does not take account of the fertilization effect
of the additional carbon dioxide (CO2) in the atmosphere, which will
increase yields of some crops. That effect has been estimated to reduce the
additional number of people at risk from hunger by 2020 to 1 million
(Parry and others 2004). Malnutrition in Latin America and the Caribbean
will be affected by droughts as well as by floods, which not only destroy
croplands but also cause vector- and waterborne diseases that increase the
risk of malnutrition by inhibiting the body’s uptake of nutrients (World
Bank 2008). Evidence from drought periods suggests that malnutrition
will increase the number of children suffering from stunted growth and
irreversibly impaired brain development; indeed young children are the
ones most at risk from malnutrition. For them the effects can be lifelong
because they are related to development, whereas adults can recover.
Malnutrition also raises the risk of mortality from other diseases, especially
malaria and diarrhea; it is the main source of mortality from both diseases
(Parry and others 2004). Malnutrition in pregnant mothers increases the
risk of subsequent infant mortality.
More frequent droughts will diminish crop production in some
parts of the LAC region, as discussed in chapter 4 (UNDP 2007). In
Mexico, crop reductions of 30 percent, the upper end of the range pro-
jected by the IPCC by 2020 (IPCC 2007), could threaten livelihoods
and food security and be exacerbated by loss of livestock. In northeast
Human Health and Climate Change 181
education: taller people got better jobs. Hence natural hazards, working
through their effects on nutrition, have the potential to affect not only
children’s current health status but also how they fare in school and per-
form in the labor market as adults.
Reduced
Erosion of
productivity
financial capital
and income
Increase in
illness and
morbidity
Erosion of human
Increased capital (health and
spending on education)
treatment and
prevention
Source: Authors.
Human Health and Climate Change 183
can cause anemia due to iron deficiency. As seen above, if gross nutri-
tional development is impeded, especially in children under the age of
two, brain development is irreversibly impaired, affecting cognitive
ability, language, attention, and behavior (Holding and Snow 2001).
Thus the consequences of ill health and malnutrition in childhood are
often long lasting, so that children (also once they become adult)
require special support from family, community, health services, and
education services.
Besides impairing brain development and cognitive abilities, malaria
can interfere with educational attainment in other ways. School atten-
dance might be interrupted by illness, or children might only be able to
attend school if it is located close to home (Holding and Snow 2001).
Although few empirical studies examine the effect that malaria can
have on educational performance, the limited evidence suggests that a
relationship exists (Chima, Goodman, and Mills 2003). Losing many
days because of absences increases the likelihood of failing and having
to repeat a grade, as well as of dropping out of school altogether (Sachs
and Malaney 2002). As for the effects on educational performance, a
study in Kenya of 87 case patients found that 3–4 years after being dis-
charged from treatment for malaria, 14 percent were performing at
least two standard deviations below the mean in school tests, compared
to only 5 percent among the control group (Holding and Snow 2001).
When treatment costs are high, a family with limited assets may decide
to stop supporting the education of some or all of its children (Holding
and Kitsao-Wekulo 2003). In addition, if parents believe that the long-
term effects of the disease will compromise their child’s educational
prospects or health status, they may decide not to invest in the child’s
education.
Malaria’s impacts on educational attainment and on labor productivity
are likely to be more significant in areas located on the edge of endemic-
ity because in those areas a smaller percentage of children will have devel-
oped immunity by the time they reach school age. (Currently, areas with
endemic malaria have low rates of school absence because school-age
children have already developed immunity [Holding and Snow 2001]).
Productivity losses from malaria are smaller in areas where the disease is
widespread, compared to areas where transmission is less common and
immunity is rare (Malaney, Spielman, and Sachs 2004).
Household income is affected by disease in several ways. Health is a
human capital asset, and its value at the household level is determined
both by the number of individuals who are able to work and by the
184 Andersen, Geary, Portner, and Verner
quality of the work they are able to do (IISD 2003). Because malaria
and malnutrition can reduce both the quantity and the quality of
household labor, they can significantly reduce income. Children who
become ill or weak might not be able to take up chores for the family,
such as working in the fields, thus reducing their contribution to family
productivity and income. Similarly, as the primary caregivers, women
may have to spend more time on child care when a child is sick, reduc-
ing their availability for activities that contribute to household food
production or income generation.
Besides lowering labor productivity, diseases such as malaria also force
households to spend on prevention and treatment, the cost of which may
erode their financial assets if they are poor (Malaney, Spielman, and Sachs
2004). Sometimes a rural household must pay hired workers to offset the
loss in productivity of its own family members (Sachs and Malaney 2002).
particularly the case for the vector-borne diseases malaria and dengue
fever, as well as for waterborne diseases. Risks are particularly serious
when diseases are brought into new areas, where immunity levels are low,
or when new virus strains are created.
Further, some of the health impacts from climate change and variabil-
ity could have wider social consequences as human capital assets are
weakened. That could hold back children’s educational performance and
reduce financial and other household assets, thereby reducing household
resilience and adaptive capacity.
To sum up, the potential health risks to the LAC region are very seri-
ous, and as the IPCC noted in its last report, adaptive capacity in the
region is low (IPCC 2007). Building up the region’s adaptive capacity and
reducing health risks resulting from climate change and variability, espe-
cially among the groups with the fewest assets, arguably are the most
important development issues facing the region today. Recommendations
for those purposes are presented below.
Notes
1. DALY refers to the sum of years of life lost due to premature death (YLL),
and years of life lived with disability (YLD).
2. Floods have caused the majority of climate change–related DALY in Latin
America, at 72,000, followed by diarrhea, at 17,000, and malaria at 2,000
(WHO 2003).
190 Andersen, Geary, Portner, and Verner
3. It has been estimated that the 2003 heat wave in Europe led to between
40,000 and 52,000 excess deaths (Larsen 2006; Sardon 2007).
4. Even greater increases are estimated by Stern (2007), who projects dengue
transmission to increase by two to five times by the 2050s.
5. This number has been put as high as 1.8 million DALY.
6. The Aedes aegypti mosquito is also known as Aedes stegomyia.
7. Anecdotal evidence from St. Lucia suggests that a public education campaign
stressing the importance of keeping water containers covered, to deny mos-
quitoes access to standing water, has successfully reduced the incidence of
vector-borne diseases.
8. This scenario envisages a mean temperature change of 3.4°C by 2090–99,
relative to 1980–99 (see IPCC 2001a).
9. See, for example, Glewwe and King 2001; Glewwe, Jacoby, and King 2001;
Alderman, Hoddinott, and Kinsey 2006; and van den Berg, Lindeboom, and
Portrait 2007. In addition there is some evidence that conditions while in
utero can affect later outcomes, as shown by Behrman and Rosenzweig 2004;
and Almond 2006.
10. As discussed in Strauss and Thomas (1995) and Wolpin (1997), the literature
on the determinants of child health and mortality in developing countries is
large. Despite this large literature, there is surprisingly little that directly deals
with the potentially adverse effects of shocks from natural hazards. Foster
(1995) showed for Bangladesh that floods lead to substantially lower weight
for the children affected and argued that credit market imperfections were
the main factor behind the differences in how children’s weights responded.
More recently, Baez and Santos (2007) examined the effects of Hurricane
Mitch in Nicaragua on children’s health, schooling, and labor force participa-
tion. They found that those affected by the hurricane were more likely to be
undernourished and that the distribution of nutritional status worsened, espe-
cially for those at the bottom of the distribution.
11. Recall that the WHO does not find any substantial evidence of different
growth patterns for children of different races exposed to “optimal” con-
ditions.
12. For each country the relationship between climate (temperature and precip-
itation) and life expectancy (or child mortality) was estimated, using munici-
pality-level data, controlling for other factors that also might affect life
expectancy (education and urbanization rates). These estimated relationships
were then used to simulate the effects of past (past 50 years) climate change
as well as expected future (future 50 years) climate change, in each munici-
pality in each of the five countries. Estimating the effects at the municipal
level has the advantage of permitting an analysis of the distributional conse-
quences of climate change on health. Effects of increased climatic variability,
Human Health and Climate Change 191
References
Acharya, A., and M. K. Paunio. 2008. Environmental Health and Child Survival:
Epidemiology, Economics, Experiences. Washington, DC: World Bank.
Alderman, H., J. Hoddinott, and B. Kinsey. 2006. “Long-Term Consequences of
Early Childhood Malnutrition.” Oxford Economic Papers 58 (3): 450–74.
Almond, D. 2006. “Is the 1918 Influenza Pandemic Over? Long-Term Effects of
In Utero Influenza Exposure in the Post–1940 U.S. Population.” Journal of
Political Economy 114 (4): 672–712.
AP (Associated Press). 2007. “Hemorrhagic Dengue Fever Surges in Mexico with
Climate Changes, Migration, Urbanization.” March 31.
Aramburú, J., C. Ramal, and R. Witzig. 1999. “Malaria Reemergence in the
Peruvian Amazon Region.” Emerging Infectious Diseases 5 (2): 209–15.
Baez, J. E., and I. V. Santos. 2007. “Children’s Vulnerability to Weather Shocks:
A Natural Disaster as a Natural Experiment.” Mimeo. Syracuse University.
http://student.maxwell.syr.edu/jebaez/Mitch&Children%27sVulnerability
(FinalVersion-Sept2007).pdf.
Behrman, J. R., and M. R. Rosenzweig. 2004. “Returns to Birthweight.” Review of
Economics and Statistics 86 (2): 586–601.
Blanco, J., and D. Hernandez. 2009. “The Costs of Climate Change in Tropical
Vector-Borne Diseases: A Case Study of Malaria and Dengue in Colombia.”
LCR Sustainable Development Working Paper 32: 69–87, World Bank,
Washington, DC.
Chen, A. 2006. The Threat of Dengue Fever in the Caribbean: Impacts and
Adaptation. Washington, DC: International START Secretariat.
Chima R. I., C. A. Goodman, and A. Mills. 2003. “The Economic Impact of Malaria
in Africa: A Critical Review of the Evidence.” Health Policy 63 (1): 17–36.
192 Andersen, Geary, Portner, and Verner
De la Torre, A., P. Fajnzylber, and J. Nash. 2008. Low Carbon, High Growth. Latin
American Responses to Climate Change. Washington, DC: World Bank.
Ebi, Kristie L. 2008. “Healthy People 2100: Modeling Population Health Impacts
of Climatic Change.” Climatic Change 88: 5–19.
ECLAC, UN, and UNICEF. 2006. “Child Malnutrition in Latin America and the
Caribbean.” Challenges 2. United Nations publication ISSN 1816-7543.
Foster, A. D. 1995. “Prices, Credit Markets, and Child Growth in Low-Income
Rural Areas.” Economic Journal 105 (430): 551–70.
Gerolomo, M., and M. L. F. Penna. 1999. “The Seventh Pandemic of Cholera in
Brazil.” Informe Epidemiologico do Sus 8 (3): 49–58.
Glewwe, P., H. G. Jacoby, and E. M. King. 2001. “Early Childhood Nutrition and
Academic Achievement: A Longitudinal Analysis.” Journal of Public Economics
81 (3): 345–68.
Glewwe, P., and E. M. King. 2001. “The Impact of Early Childhood Nutritional
Status on Cognitive Development: Does the Timing of Malnutrition Matter?”
World Bank Economic Review 15 (1): 81–113.
Gubler, Duane J. 2002. “Epidemic Dengue/DHF as a Public Health, Social, and
Economic Problem in the 21st Century.” Trends in Microbiology 10 (2): 100–03.
Holding, P., and P. Kitsao-Wekulo. 2003. New Perspectives on the Causes and
Potential Costs of Malaria. Disease Control Priorities Project Working Paper
No 7, World Bank, Washington, DC. http://www.dcp2.org/file/23/WP7.pdf.
Holding, P., and R. Snow. 2001. “Impact of Plasmodium Falciparum Malaria on
Performance and Learning: Review of the Evidence.” American Journal of
Tropical Medicine 64 (1): 68–75.
IISD (International Institute for Sustainable Development). 2003. Livelihoods and
Climate Change. Winnipeg, Canada: IISD.
IPCC (Intergovernmental Panel on Climate Change). 2001a. Special Report on
Emission Scenarios. Geneva: IPCC.
———. 2001b. Third Assessment Report, Working Group II. Geneva: IPCC.
———. 2007. Climate Change 2007: Impacts, Adaptation, and Vulnerability.
Contribution of Working Group II to the Fourth Assessment Report of the
IPCC. Geneva: IPCC.
Iwanciw, Javier Gonzales, Jorge Cusicanqui Giles, and Marilyn Aparicio Effen, eds.
2004. Vulnerabilidad y Adaptación al Cambio Climático en las Regiones del
Lago Titicaca y los Valles Cruceños de Bolivia. La Paz: Government of Bolivia,
Ministry of Development Planning.
Larsen, Janet. 2006. “Setting the Record Straight: More than 52,000 Europeans
Died from Heat in Summer 2003.” Earth Policy Institute. Web article, July 28,
2006. http://www.earth-policy.org/Updates/2006/Update56.htm.
Human Health and Climate Change 193
Maccini, S., and D. Yang. 2008. “Under the Weather: Health, Schooling, and
Socioeconomic Consequences of Early-Life Rainfall.” Mimeo. University of
Michigan.
Magrin, G., C. Gay García, D. Cruz Choque, J. C. Giménez, A. R. Moreno,
G. J. Nagy, C. Nobre, and A. Villamizar. 2007. Latin America. Climate Change
2007: Impacts, Adaptation and Vulnerability. Contribution of Working Group II
to the Fourth Assessment Report of the Intergovernmental Panel on Climate
Change, ed. M. L. Parry, O. F. Canziani, J. P. Palutikof, P. J. van der Linden and
C. E. Hanson. 581–615. Cambridge, U.K.: Cambridge University Press.
Malaney, P., A. Spielman, and J. Sachs. 2004. “The Malaria Gap.” American Journal
of Tropical Medicine 71 (suppl. 2): 141–46.
Mata, L., and C. Nobre. 2006. “Impacts, Vulnerability, and Adaptation to Climate
Change in Latin America.” Background paper commissioned by UN Framework
Convention on Climate Change, Lima, Peru.
Moreno, Ana Rosa. 2006. “Climate Change and Human Health in Latin America:
Drivers, Effects, and Policies.” Regional Environmental Change 6: 157–64.
PAHO (Pan-American Health Organization). 1995–2006. Numbers of Reported
Cases of Dengue and Malaria.Washington DC: PAHO.
———. 2008. Numbers of Reported Cases of Dengue and Dengue Hemorrhagic Fever
(DHF) in the Americas, by Country: Figures for 2008. Washington DC: PAHO.
Parry, M. L., C. Rosenzweig, A. Iglesias, M. Livermore, and G. Fischer. 2004. “Effects
of Climate Change on Global Food Production under the SRES Emissions and
Socioeconomic Scenarios.” Global Environmental Change 14: 53–67.
Patz, Jonathan A., and John M. Balbus. 1996. “Methods for Assessing Public
Health Vulnerability to Global Climate Change.” Climate Research 6: 113–25.
Patz, Jonathan A., and R. Sari Kovats. 2002. “Hotspots in Climate Change and
Human Health.” BMJ 325: 1094–98.
Sachs, Jeffrey, and Pia Malaney. 2002. “The Economic and Social Burden of
Malaria.” Nature 415: 113–25.
Sardon, J. P. 2007. “The 2003 Heat Wave.” Eurosurveillance 12, no. 3 (March): 694.
01 European Centre for Disease Prevention and Control, Stockholm.
http://www.eurosurveillance.org/ViewArticle.aspx?ArticleId=694.
Stern, Nicholas. 2007. Stern Review on the Economics of Climate Change. Cambridge,
U.K.: Cambridge University Press.
Strauss, J., and D. Thomas. 1995. “Human Resources: Empirical Modeling of
Household and Family Decisions.” In Handbook of Development Economics, vol.
3A, ed. J. Behrman and T. N. Srinivasan, 1883–2023. Amsterdam, New York,
and Oxford: Elsevier Science.
———. 1998. “Health, Nutrition, and Economic Development.” Journal of Economic
Literature 36 (2): 766–817.
194 Andersen, Geary, Portner, and Verner
Torres, Jaime, and Julio Castro. 2007. “The Health and Economic Impact of
Dengue in Latin America.” Cadernos de Saúde Pública 23 (suppl. 1): 23–31.
Rio de Janeiro. doi: 10.1590.
UNDP (United Nations Development Program). 2007. Fighting Climate Change:
Human Solidarity in a Divided World, Human Development Report 2007/08.
New York: Oxford University Press.
Van den Berg, G. J., M. Lindeboom, and F. Portrait. 2007. Long-Run Longevity
Effects of a Nutritional Shock Early in Life: The Dutch Potato Famine of
1846–1847. Discussion Paper 3123, Institute for the Study of Labor (IZA),
Bonn, Germany.
Warren, R., N. Arnell, R. Nicholls, P. Levy, and J. Price. 2006. Understanding the
Regional Impacts of Climate Change. Tyndall Center for Climate Change
Research Working Paper 90, Norwich, U.K.
Wolpin, K. I. 1997. “Determinants and Consequences of the Mortality and Health
of Infants and Children.” In Handbook of Population and Family Economics, vol.
1A, ed. M. R. Rosenzweig and O. Stark, 483–557. Amsterdam: Elsevier Science.
WHO (World Health Organization). 2003. Climate Change and Human Health.
Geneva: WHO.
———. 2005. Using Climate to Predict Infectious Disease Epidemics. Geneva:
WHO.
———. 2008. Dengue and Dengue Hemorrhagic Fever. Fact Sheet No. 117. Geneva:
WHO.
World Bank. 2008. Environmental Health and Child Survival: Epidemiology,
Economics, Experiences. Washington, DC: World Bank.
CHAPTER 7
195
Table 7.1 Hot Spots for Selected Climate Change Events with Potential to Spur Migration
Hot spots with
196
Migration Patterns
Migration, internal as well as intraregional and international, plays an
important role in Latin America and the Caribbean, which had the largest
number of migrants of any developing region in 2000–05 (World Bank
2008). As of 2005, some 25 million Latin American and Caribbean
migrants were abroad, representing more than 13 percent of migrants
worldwide (ECLAC 2006) and about 4 percent of the region’s total popu-
lation of 559 million.
Since the 1950s the net emigration rate from the LAC region has
steadily risen, but it is expected to peak at approximately 1.1 million per
year between 2000 and 2010. Subsequently it is projected to fall to close
to 600,000 per year by 2020 and to stabilize at that level until 2050, which
is the horizon for the current forecasts (figure 7.1).
Migrant remittances to LAC countries have been growing steadily, in
absolute terms (figure 7.2) and as a share of the region’s gross domestic
product (GDP), reaching 2 percent in 2006 (ECLAC 2006). By 2006,
remittances to LAC countries, at US$68 billion, were surpassed only by
those into Asian countries, at US$114 billion (IADB 2007). However, the
economic crisis that gripped the world in 2008 led remittances to Latin
America and the Caribbean to remain stagnant in the third quarter of
2008, and the fourth quarter saw a decline for the first time since track-
ing of them began in 2000 (IADB 2009).
Figure 7.1 Net Emigration, Latin America and the Caribbean, 1950–2050
2,000
1,800
1,600
1,400
thousands
1,200
1,000
800
600
400
200
0
0
0
–6
–7
–8
–9
00
–1
–2
–3
–4
–5
50
60
70
80
00
10
20
30
40
–2
19
19
19
19
90
20
20
20
20
20
19
Figure 7.2 Growth in Remittances Received in Latin America and the Caribbean,
1997–2006
60
50
40
US$ billions
30
20
10
0
1997 1998 1999 2000 2001 2002 2003 2004 2005 2006
Source: Authors’ elaborations based on data from World Bank, World Development Indicators, various years.
10
9
remittances received averaged over
7
1970–2007 (% GDP)
0
0 1 2 3 4 5 6
accumulated loss from natural disasters 1997–2007 logged
(% 2007 GDP)
500
GDP, indexed (1995 =100)
400
300
200
100
0
1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005
Source: Authors’ elaboration.
Note: It has been difficult to find countries where there is a discernible effect on remittances. In the case of
Nicaragua, one could say that an increase can be observed at the time of Hurricane Mitch in 1998. The remit-
tances (as percentage of GDP) shown on the y-axis have been indexed (1995 = 100). The explosive increase in
remittances covers up small variations caused by natural disasters, many of which are local. Thus even if local
populations are very dependent on remittances it might not show up in aggregate statistics.
Because land is not a mobile asset, landowners are to a large extent tied
down by their land, especially if they do not have a title and therefore
cannot easily sell the land at an acceptable price. Landowners and their
families seem to be willing to stay in place longer than households with
limited access to land and other physical assets.
Most of the region’s migrants are young men, but the region now also
has one of the world’s highest migration rates for women (Deshingkar
and Grimm 2004). In some Latin American countries, access to land is so
limited that nearly all the poor young people in rural areas view migra-
tion as their main, and perhaps only, livelihood option. Across the region
during 1988–2003, 15.5 million rural young people (aged 15–29) became
part of the urban population. This translates into a reduction of one-third
of the rural population of this age group (table 7.2).
areas. The plan makes it a condition for the farmers who participate
that they abandon their traditional crops, such as corn and beans, and
replace them with crops that are in high demand in international mar-
kets, such as oranges and palm oil. The productive conversion should
increase the households’ income, as well as their knowledge regarding
sustainable land use. However, abandoning staple crops completely in
favor of cash crops can leave the farmers vulnerable to volatility in
world markets. Families are expected to benefit from improved health
services, electricity, information technology, schools, child care, and
other services, which were much less available in their old settlements.
More general experience with resettlement schemes, for example,
those associated with big dam projects, shows that when done properly
and inclusively, planned population movement can avoid the danger of
sparking potential conflict with and within the communities. Still, a
risk exists that the households and their members may lose their tradi-
tions and spiritual connection to their land. This is a risk especially for
indigenous peoples (box 7.1).
Box 7.1
Table 7.3 Bolivia: Regression Findings for the Three Migration Models
Model 2:
Model 1: Destination Model 3:
Explanatory variables Out-migration choice Migration rates
Constant 105.385 12.845 1.484
(6.01) (13.65) (13.76)
Distanceij –14.771 –1.687
(–47.73) (–51.42)
Characteristics of sending municipality
Average temperaturei –10.163 –0.107
(–4.21) (–12.48)
(Average temperaturei)2 0.401 0.004
(4.73) (14.26)
Average rainfalli –51.864 –0.175
(–3.15) (–7.62)
(Average rainfalli)2 12.720
(2.59)
Per capita consumptioni 0.268 0.002
(2.97) (5.34)
(Per capita consumptioni)2 –0.0004 –0.000003
(–2.21) (–4.39)
Share in secondary and tertiary –0.370
sectorsi (–2.42)
(Share in secondary and tertiary 55.2970 0.756
sectorsi)2 (3.67) (4.59)
Urbanization ratei –0.197
(–4.12)
Education leveli
Municipal spending per capitai –0.171 0.001
(–2.24) (4.98)
Population densityi –0.410 –0.003
(–2.92) (–5.45)
Share of municipality with 0.512 0.002
steep slopei (5.09) (6.76)
Density of secondary roadsi
Oil concessioni –0.066
(–3.12)
Forestry concessioni 0.110
(4.66)
Characteristics of receiving municipality
Average temperaturej 11.105 –0.119
(–11.00) (–12.98)
(Average temperaturej)2 0.045 0.005
(12.76) (15.11)
(continued)
210 Andersen, Lund, and Verner
Table 7.3 Bolivia: Regression Findings for the Three Migration Models (continued)
Model 2:
Model 1: Destination Model 3:
Explanatory variables Out-migration choice Migration rates
Average rainfallj –1.9608 –0.1497
(–2.99) (–6.32)
(Average rainfallj)2 0.3831
(2.00)
Population sizej 0.9955 0.0870
(84.26) (78.59)
Per capita consumptionj -0.0255 –0.0021
(–12.17) (–11.45)
Education levelj 0.4781 0.0472
(6.04) (6.57)
Municipal spending per capitaj 0.0022
(8.09)
Share in secondary and tertiary 6.7452 0.5533
sectorsj (10.42) (9.21)
Pressureja –0.9509 –0.0751
(–5.31) (–4.53)
Municipality areaj –0.0382 –0.0028
(–3.17) (–2.47)
Population densityj 0.1710 0.0107
(28.12) (19.20)
Urbanization ratej
Density of primary roadsj
Density of secondary roadsj –0.0072 –0.0005
(–5.33) (–4.23)
Share of municipality with 0.0130 0.0013
steep slopej (3.41) (4.03)
Oil concessionj
Forest concessionj 0.8824 0.0877
(3.43) (3.68)
Number of observations 312 98282 97032
R2 0.2648 0.2172 0.1959
Source: Authors’ estimation.
Notes: The subscript i’s in the stub column refer to sending municipality information; the subscript j’s refer to
receiving municipality information; t-values in parentheses.
a. Pressure is the ratio of population to total employment: an indication of the level of unemployment or lack of
economic opportunities. It is expected to be positively related to out-migration rates and negatively related to
in-migration rates.
158
108
58
8
3 7 11 15 19 23 27
average annual temperature (ºC)
158
108
58
8
0 1 2 3
average accumulated precipitation (m)
Source: Authors’ elaboration.
Note: The green central line represents the point estimate from Model 1 of table 7.3. The thinner black lines
delimit the 95 percent confidence interval as estimated by Stata’s lincom command.
simply too poor to shoulder the costs of migrating. Finally, the remaining
coefficients indicate that people are abandoning the most sparsely popu-
lated municipalities and the most rugged terrain, but municipal spending
can help to keep them in place.
8
additional migrants per 1,000 inhabitants
–2
–4
–6
–8
–10
200 1,200 2,200
consumption per capita 2001
(ppp-adjusted 2001 US$)
change; in some of the very poorest, the rate is 6–8 persons per 1,000
inhabitants every 5 years.6
section takes advantage of a migration intensity index that has been esti-
mated for each of 2,350 municipalities in Mexico by the National
Population Council of Mexico, using information from the 2000 popula-
tion census (Consejo Nacional de Población 2002). The index integrates
information about current migrants in the United States, migrants recently
returned from the United States, and remittances received. The explana-
tory variables of principal interest are average annual temperature and
annual rainfall, and as control variables, the literacy rate and urbanization
rate in each municipality have been included. The calculations allow for
nonlinear effects for all four variables, by including the square of each.
The regression results indicate that temperature is one of the most
important variables explaining migration intensities in Mexico (table 7.5).
It appears to be even more important than education, and much more
important than income.
Figure 7.8 illustrates the estimated nonlinear relationships between cli-
mate and migration intensities. Contrary to expectations, migration inten-
sities are maximized at moderate temperatures (around 20°C), even
when controlling for other factors that also might affect migration rates.
Migration rates are also significantly related to average annual rainfall,
being highest in municipalities with very little rain. This implies that a
reduction in rainfall would tend to increase migration to the United
States, whereas a general temperature increase would have ambiguous
effects. It is therefore necessary to do a detailed, municipality-level simu-
lation to see which effects dominate.
Figure 7.8 Estimated Relationships between Temperature and Rainfall and Migration Intensity
4 4
3 3
migration intensity index
1 1
0 0
–1 –1
–2
8 12 16 20 24 28 –2
0 1 2 3 4 5 6
average annual temperature (ºC)
average accumulated precipitation (m)
Notes
1. The definition used here is that put forward by the International Organization
for Migration (IOM): “Persons or groups of persons who, for compelling rea-
sons of sudden or progressive changes in the environment that adversely
affect their lives or living conditions, are obliged to leave their habitual homes,
or choose to do so, either temporarily or permanently, and who move either
within their country or abroad.”
2. http://en.wikipedia.org/wiki/Hurricane_Mitch.
218 Andersen, Lund, and Verner
3. Note that the climate data used relate to average monthly temperatures and
precipitation levels; increases in climate variability and severe weather events
are not included, except to the extent that they are related to average tem-
perature and precipitation. Specific effects of changes in glacier melt are not
included either.
4. This was done by using the stepwise function in Stata.
5. This is driven by relatively large drops in nighttime temperatures, as summer-
and daytime temperatures have risen.
6. In another municipality-based study of migration, in this case for Brazil,
Assunção and Feres (2008b) found that high temperatures and low precipita-
tion levels at the origin act as push factors, while lower temperatures and
more precipitation at the destination act as pull factors. They also found the
greatest effect when the distanced traveled is greater than 100 kilometers; for
migrations of less than 100 kilometers from the origin hardly any climate
effect is detectable.
7. For this reason it would be preferable to work with raw migration rates
instead of composite indexes.
References
Altamirano, T., R. Hirabayashi, and X. Albó. 1997. Migrants, Regional Identities,
and Latin American Cities. Society for Latin American Anthropology
Publication Series 13. Arlington, VA: American Anthropological Association.
Assunção, J., and F. C. Feres. 2008a. “Climate Change, Agricultural Productivity,
and Poverty.” Background paper for De la Torre, A., P. Fajnzylber, and J. Nash
(2009), Low Carbon, High Growth—Latin American Responses to Climate
Change: An Overview.” Washington, DC: World Bank.
———. 2008b. “Climate Migration.” Background paper for De la Torre, A.,
P. Fajnzylber, and J. Nash (2009), Low Carbon, High Growth—Latin American
Responses to Climate Change: An Overview. Washington, DC: World Bank.
Belcher, J. C., and F. L. Bates. 1983. “Aftermath of Natural Disasters: Coping
through Residential Mobility.” Demography 7.
Bravo, J. 2008. “Migration, Population Distribution, and Development.” Presentation
at Expert Group Meeting on Urbanization, Internal Migration, Population
Distribution, and Development, New York, January 21–23, United Nations
Population Division.
Brown, Oli. 2008. Migration and Climate Change. IOM Migration Research Series
31. Geneva: International Organization for Migration.
CEPAL (Comisión Económica para América Latina y el Caribe). 2006. Panorama
Social de América Latina 2006. Publicaciones de las Naciones Unidas.
Migration and Climate Change 219
221
222 Rubin
Empirical studies provide only limited support for a link between cli-
mate change and an increasing risk of conflict. Whereas most research on
climate change and variability is published in respected, peer-reviewed
journals, the research that relates climate change to conflict consists largely
of speculative scenarios published on the Web sites of political organiza-
tions and think tanks.2 That is cause for concern.
Conflict and Climate Change 225
violent conflicts. These authors report that the strongest “result is that the
ecological footprint, perhaps the most widely accepted measure of man’s
use (or abuse) of ecoservices, is positively correlated with peace, a result
that is stubborn and substantively large.” A recent assessment of the
empirical literature on climate change and conflict (Buhaug, Gleditsch,
and Theisen 2008) highlights the lack of robustness by concluding that
several of the earlier findings are either not replicable or do not hold with
improved data.
Evidence indicates that climate change will have only a negligible
impact on the risk of armed or violent conflict. The conclusions from the
literature, combined with the LAC region’s limited propensity for violent
conflict, suggest that climate change will not significantly influence the
risk of conflict. Even so, such an inference seems problematic. The effects
of climate change and variability work through multiple and often inter-
related channels, some of which have not yet received enough academic
scrutiny. Whereas the effects of climate change are local, dynamic, and
multidimensional, the existing studies rely strongly on aggregate-level
effects of environmental scarcity on conflict. And whereas the focus has
traditionally been on violent conflicts, a need also exists for analysis that
encompasses nonviolent conflicts. The latter are more numerous, and the
way in which they are resolved (inclusively and equitably, or by means
that perpetuate tendencies of coercion, exclusion, inequality, and stigma-
tization) affects social development outcomes.
In short, it seems premature to dismiss the potential threats that cli-
mate change and variability pose to national security and development. It
is noteworthy that from a social development perspective many of the
existing studies rest on misspecifications of the two variables of interest—
climate change (the explanatory variable) and conflict (the dependent
variable)—and thus risk seriously underestimating the impact of climate
change on conflict in Latin America and the Caribbean. In the following
section these two variables are recalibrated from a social perspective to
obtain better approximations of the extent and nature of climate
change–induced conflicts in the LAC region.
caused by climate change; (d) the cumulative and threshold effects of cli-
mate change; and (e) the prospects for using climate change as a cause for
political mobilization.
The adverse effect of migration on the risk of conflict has been well
documented. Although Urdal (2005) did not find evidence of a link
between migration and conflict, both Salehyan and Gleditsch (2006) and
Reuveny (2007, 2008) report that migration appears to increase the risk
of conflict.6
Conflict and Climate Change 229
water conflict has been triggered by the U.S. plan to reline part of the All-
American Canal. Such a project would minimize leakage for California
users but in the process would impose water shortages on villagers in
Mexico. The plan has spurred organized protests from Mexico, with the
president noting that it would be “terrible for Mexico.”
Socioeconomic Context
Gaining an overall view of the socioeconomic conditions in Latin America
and the Caribbean that could exacerbate the conflictual dimension of cli-
mate change is essential. For example, the region’s high income inequal-
ity and crime rates provide a backdrop of tension to which the adverse
impacts of climate change and variability are likely to add further stress
that could increase the risk of violent and nonviolent conflicts.
Compared to African countries, countries in the LAC region in gen-
eral are relatively strong, middle-income societies with accountable
Conflict and Climate Change 235
Table 8.3 Share of Countries in Different Regime Categories and Regime Stability
across Categories—Comparison of LAC Region and the World
Political
rights
regime
score 1 2 3 4 5 6 7 Average Median Variance SD
LAC 27% 29% 13% 10% 8% 7% 6% 2.9 2.0 1.5 1.0
World 23% 15% 7% 9% 11% 19% 16% 3.9 4.0 1.2 0.9
Bolivia – – – – – – – 2.9 – 3.1 –
Colombia – – – – – – – 2.7 – 0.8 –
Mexico – – – – – – – 3.4 – 0.7 –
Nicaragua – – – – – – – 4.2 – 1.1 –
Peru – – – – – – – 3.8 – 3.4 –
Source: Author’s calculations based on Freedom House data.
Note: The numbers heading the columns indicate the seven Freedom House regime categories, ranging from
1 = most democratic, through 7 = least democratic. Region and selected LAC countries 1973–2006; unweighted
averages. Regime stability is proxied by regime category variance and standard deviation within countries and
averaged across countries. — = not available.
rule (32, 30, and 28 years respectively). Chile, not surprisingly, is the coun-
try with the greatest change in rule, from military rule to the present
democratically elected president.
Haldén (2007) notes that although countries in the LAC region are not
as poorly integrated economically as those in Africa, “they are nonetheless
fraught with tensions and huge disparities in standards of living and
Conflict and Climate Change 237
income.” This, he argues, is worrisome for three reasons: (a) disparities and
divisions might by themselves impede growth and undermine adaptation
strategies; (b) the substantial inequality might also destabilize societies and
increase the risk of conflict in the light of climate change and variability;
and (c) the differences between large segments of the populations also
imply that climate change will have very unequal impacts on the popula-
tion, further exacerbating tensions.
Large inequalities between groups (along ethnic, religious, political,
or geographical divisions) increase the risk of violent conflict, whereas
high individual income inequality is a driver of crime (Dahlberg and
Gustavsson 2005; Fajnzylber, Lederman, and Loayza 2002; Kelly 2000;
Østby 2007). Violent conflicts entail coordinated group action; as Østby
argues, “Even though an individual may feel frustrated if he is poor com-
pared to other individuals in society, he will not start a rebellion on his
own.”13 Based on quantitative analysis of survey data from 55 countries,
Østby (2007) found that all measures of horizontal inequality—ethnic,
religious, political, and geographic—and in particular geographic hori-
zontal inequality, are associated with higher risks of conflict outbreaks. In
Bolivia, highly unequal distributions of natural resources create tensions
between regions. In response to the nationalization of gas and oil
reserves, the resource-rich regions of Santa Cruz, Tarija, Beni, and Pando
(comprising 35 percent of the Bolivian population) declared autonomy.
Violent disputes erupted between the government and the regions seek-
ing autonomy, and also within those regions between supporters of the
president and separatists.
The impact of income inequality on crime rates has also been ana-
lyzed. The high crime rates in Latin America and the Caribbean are likely
rooted partly rooted in the region’s high level of income inequality.
Gutierrez and others (2004) concluded, based on their econometric
analysis of homicide rates in Brazil, that individual inequality (as meas-
ured by the Gini coefficient), but not poverty, played a substantial role in
determining criminality. LAC countries in general, and countries such as
Brazil, Colombia, and Haiti in particular, are among the most unequal in
the world (table 8.4). Climate change and variability might raise crime
rates by causing events (natural disasters, sea level rises, and ecological
degradation) that worsen inequality. Adverse exogenous shocks to soci-
eties are rarely absorbed equally by different income strata.
Higher crime rates could lead to a weakening of state capacity. Crime
of most types has been rising in Latin America and the Caribbean since
1980, mirroring global trends. In 1980, the LAC region had a little more
238 Rubin
Table 8.4 Income Inequality across Regions and within LAC Countries
(Gini coefficients)
LAC 50.7
Haiti 59.2
Brazil 59.0
Colombia 56.4
Bolivia 54.0
Mexico 49.7
Argentina 49.5
Peru 48.8
Guyana 47.4
Costa Rica 46.4
Nicaragua 46.3
Trinidad and Tobago 40.7
Sub-Saharan Africa 50.1
East Asia 39.0
Middle East 38.7
North America 36.7
South Asia 34.9
Europe and Central Asia 31.9
Source: World Development Indicators (World Bank 2009).
Note: Unweighted averages 1950–2006; based on data from 131 countries where data were available.
Policy Perspectives
Climate change does not seem to call for an immediate change in policies
or for specific programs in the area of conflict prevention for the time
being. With respect to conflict, the focus should be on considering the
political economy dimension because the political economy that (a) dic-
tates the policies that can be implemented to address the adverse socioe-
conomic consequences of climate change; (b) produces the structures
that reproduce and perpetuate adverse conflictual or nonconflictual out-
comes from climate change; and (c) influences the risk that existing
socioeconomic factors become conduits for violent conflict.
Because multiple channels exist through which climate change might
affect conflict, policies need to address the consequences of climate
change on several fronts. Doing that will reduce the risk of violent con-
flict and ensure that conflicts are resolved in ways that do not jeopardize
key development objectives. Policies can either seek to prevent the
adverse socioeconomic consequences of climate change from turning into
242 Rubin
Notes
1. Based on conflict data from the Peace Research Institute, Oslo (1946–2006),
the LAC average of one conflict per 23,714 inhabitants is higher than that in
Asia (one conflict per 57,382 inhabitants) and about the same as in Europe
(one conflict per 22,250 inhabitants), but significantly lower than in the
Middle East (one conflict per 12,459 inhabitants) or Africa (one conflict per
10,161 inhabitants). Most LAC countries are middle income countries, and
while they account for 29 percent of the middle-income countries globally,
they are only involved in 19 percent of the middle-income-country conflicts
(PRIO 2009).
Conflict and Climate Change 243
2. With the award of the 2007 Nobel Peace Prize to Al Gore and the
Intergovernmental Panel on Climate Change (IPCC), however, the conflictual
consequences of climate change and variability are likely to receive increased
academic attention.
3. Despite their clear conviction that population growth and environmental
degradation do influence conflict, Homer-Dixon and Percival (1995)
acknowledge that “environmental scarcity is always enmeshed in a web of
social, political, and economic factors, and its contribution to violence is
exceedingly difficult to disentangle from contributions by these other factors.”
4. See Environment, Population, and Security (1994–96), a joint effort with
the American Association for the Advancement of Science, http://www
.library.utoronto.ca/pcs/eps.htm; and Environmental Scarcity, State
Capacity, and Civil Violence (1994–98), undertaken with the American
Academy of Arts and Sciences, http://www.library.utoronto.ca/pcs/state
.htm, both under the supervision of Homer-Dixon. Under the Environment
and Conflicts Project (ENCOP, http://www.isn.ethz.ch/isn/Digital-Library/
Publications/Detail/?lng=en&id=235), 60 academic experts analyzed some
40 environmental conflicts. The Inventory of Conflict and Environment (ICE),
http://american.edu/TED/ice/ice.htm, under the supervision of Prof. Jim Lee
at the American University, contains more than 200 case studies based on
16 categories where some kind of interaction has been identified between
conflict on one side and environmental factors on the other.
5. One of the few systematic studies that examines the link between environmen-
tal scarcity on one hand and both conflict and cooperation on the other is Aaron
Wolf’s (2007) historical dataset of countries’ responses, whether cooperative or
conflictual, to water scarcity. Of the 1,831 identified interactions based on water
scarcity, 67 percent relied on cooperation, 28 percent displayed a conflictual
nature, and only 2 percent involved some type of military intervention.
6. These studies focus on the conflict impact of migration in the recipient coun-
try. To estimate the overall impact of migration on the risk of conflict, it would
be prudent to also include the effect of migration in the country of origin.
7. Despite the clear conclusion that opportunities matter in conflict, it is the
grievance angle that has been pursued in most of the empirical analyses. And
it is somewhat ironic that some of the studies most relevant for the debate
about climate change and variability and conflict are devoid of environmen-
tal indicators.
8. Diamond (2005) provides anecdotal historical evidence of societies that
collapsed because ecological degradation apparently crossed a tipping point
for civilization survival—Easter Island, Pitcairn Island, Henderson Island, the
Native American societies of Anasazi and the Maya, and Norse Greenland. In
very isolated island states the response was more-or-less apathetic; in other
cases, as with the Vikings, the response was expansionist.
244 Rubin
References
Barnett, J., and Adger, W. Neil. 2007. “Climate Change, Human Security, and
Violent Conflict.” Political Geography 26 (6): 639–55.
Binningsbø, H., I. Soysa, and N. Gleditsch. 2006. “Green Giant or Straw Man?
Environmental Pressure and Civil Conflict, 1961–99.” Working Paper presented
Conflict and Climate Change 245
PRIO (Peace Research Institute, Oslo) Armed Conflict Dataset. 2009. Version 4.
http://www.prio.no/CSCW/Datasets/Armed-Conflict/UCDP-PRIO/.
Raleigh, C., and H. Urdal. 2007. “Climate Change, Environmental Degradation,
and Armed Conflict: Disaggregated Dataset on Conflict and Environment.”
Political Geography 26 (6): 674–94.
Reuveny, R. 2007. “Climate Change–Induced Migration and Violent Conflict.”
Political Geography 26 (6): 656–73.
———. 2008. “Ecomigration and Violent Conflict: Case Studies and Public Policy
Implications.” Human Ecology 36 (1): 1–13.
Salehyan, Idean, and Kristian Skrede Gleditsch. 2006. “Refugees and the Spread
of Civil War.” International Organization 60 (2): 335–66.
Shaw, M., J. Dijik, and W. Rhomberg. 2003. “Determining Trends in Global Crime
and Justice: An Overview of Results from the United Nations Survey of
Crime Trends and Operations of Criminal Justice Systems.” Forum on Crime
and Society 3 (1): 35–62.
Sheridan, Barrett. 2008. “The People’s Pugilist.” Newsweek, April 22. http://www
.newsweek.com/id/133279?tid=relatedcl.
United Nations General Assembly. 2008. “General Assembly’s debate testament
to the importance of taking immediate practical steps to address climate
change threat, says Assembly President. . . .” GA/10690. http://www.un.org/
News/Press/docs/2008/ga10690.doc.htm.
United Nations Population Division. 2006. Data on Youth in Latin America.
http://esa.un.org/unpp/index.asp.
UNODC (United Nations Office of Drugs and Crime). 2008. United Nations
Surveys on Crime Trends and the Operations of Criminal Justice Systems.
http://www.unodc.org/unodc/en/data-and-analysis/United-Nations-Surveys-
on-Crime-Trends-and-the-Operations-of-Criminal-Justice-Systems.html.
Urdal, H. 2005. “People vs. Malthus: Population Pressure, Environmental
Degradation, and Armed Conflict Revisited.” Journal of Peace Research 42
(4): 417–34.
Urdal, Henrik. 2008. “Population, Resources and Political Violence: A Sub-National
Study of India 1956–2002.” Submitted to Development and Change. http://
www.wilsoncenter.org/events/docs/Urdal_2.pdf.
Wolf, Aaron. 2007. “Shared Waters: Conflict and Cooperation.” Annual Review of
Environment and Resources (32): 241–69.
World Bank. 2009. World Development Indicators. Washington, DC: World Bank.
http://go.worldbank.org/6HAYAHG8H0.
Yardley, Jim. 2007. “Brazil’s Leader Speaks Out.” New York Times, February 7.
http://www.nytimes.com/2007/02/07/world/asia/07china.html?page
wanted=all.
CHAPTER 9
Human beings are very versatile and inhabit both extremely cold places
(for example, the Inuit people of the Arctic region) and extremely hot
ones (for example, the Berber people of North Africa, where air temper-
atures often exceed 50°C). The Bedouins of the Sahara inhabit a place so
extremely dry that it may go for years without a drop of rain. In the
extremely wet environment where the Emberá people of Lloró,
Colombia, live it usually rains more than 13 meters per year. All other
things being equal, however, human development is easier to achieve in
places with intermediate climates, and such areas tend to support much
higher population densities and more prosperous societies than places
with extreme climates.
If temperate climates are more conducive to human development than
ones that are either too cold or too hot, then the long-run relationship
between temperature and income levels must look something like the
one depicted in Figure 9.1, which simply indicates that countries with
moderate climates are richer than countries with extreme climates.
If such an inverted-U-shaped relationship between temperature and
incomes exists, then poor countries will be more sensitive to climate
change than rich countries, simply because the slope of the relationship
is steeper at the extremes. A small change in temperatures for countries
249
250 Andersen and Verner
high
rich
income
po
or
po
or
low
too cold climate too hot
Source: Authors.
in the optimal range, where the slope is relatively flat, would cause only
a minor change in incomes, whereas a small change at the extremes
would cause large changes in incomes.
Apart from this general relationship between average climate and aver-
age income, other good arguments exist to suggest that the poor face more
serious consequences of climate change and variability than the rich. For
example, the rural poor are more likely to depend on agriculture, an activ-
ity whose productivity fluctuates strongly in response to variations in rain-
fall and temperatures. The poor are also more likely to live on marginal
lands, which are ecologically fragile and vulnerable to droughts, floods,
mudslides, and other natural disasters. The poor are less likely to be served
by public infrastructure, such as drainage canals, that can reduce the prob-
ability of weather events’ turning into natural disasters. In addition, the
poor have smaller savings to cope with adverse shocks, which means that
they may have to sell productive assets, take children out of school,
reduce nutrient intake, or take other measures that have adverse long-run
effects on the well-being of the household. A related issue is poor peo-
ple’s inability to insure themselves against adverse shocks, as well as the
generally lower level of social insurance in poor countries. If climate
change and variability hurt the poor more than the rich, it would imply
that poverty and inequality would increase as a consequence of climate
change and variability.
There is some evidence to back up this hypothesis. A recent cross-
country study by Dell, Jones, and Olken (2008) suggests that income
growth in poor countries has been very sensitive to the climate change
experienced over the last 50 years, whereas rich countries seem to
Simulating the Effects of Climate Change on Poverty and Inequality 251
have been unaffected. However, that study does not control for other
factors that may have affected growth, such as investment rates, educa-
tion levels, war, and AIDS, for example, which implies that differences in
all these factors are automatically attributed to climate change. To bet-
ter control for historical differences that are likely unrelated to future
climate change, Horowitz (2006) suggests using data on within-country
variation in incomes and climate gathered from large, heterogeneous
countries. That approach is used in this chapter. The approach makes it
possible to bring relatively detailed, local-level data to bear on the rela-
tionship between climate and income in individual countries. Though it
is always risky to draw inferences about the future from cross-sectional
relationships, these relationships are used to assess the likely direction
and magnitude of the effects of anticipated future climate change on
poverty and income inequality. In this chapter no attempt is made to
estimate the effects of changes in climate variability or changes in the
frequency of extreme events, as data to quantify such changes at the
local level are not available.
The findings of the analysis show that the impacts of climate change are
very situation specific, so that caution should be used before generalizing.
For example, in Brazil poverty is more widespread in the hot north and
northeast than in the south and southeast. Northern Brazil is expected to
warm faster than southern Brazil, and people living in already hot areas
are expected to suffer more from additional warming than people in
cooler areas. The combination of these three factors implies that the pro-
jected warming in Brazil would tend to cause an increase in both poverty
and inequality. In neighboring Bolivia, however, the situation is inverted.
The poor are concentrated in the cold highlands and would therefore
benefit more from additional warming than the wealthier population in
the hot lowlands. This means that future warming could tend to reduce
both poverty and inequality in Bolivia.
Chile on 324 comunas; and in Peru on 1,829 distritos. For simplicity, all
these local administrative units are called “municipalities” in the remainder
of the chapter.
To estimate a short-run, contemporary relationship between climate
and income, rather than a long-run, historical relationship, the study con-
trols for variables that are clearly related to income, but which are prob-
ably not affected by climate change in the short run (that is, within a few
decades). For example, although climate change is likely to affect agricul-
tural productivity in the short run, it is unlikely to reduce education lev-
els or reverse the urbanization process, except possibly in the very long
run. Therefore both education and urbanization rates are included as con-
trol variables in the short-run regressions, as outlined in appendix B. The
regressions also include rainfall and rainfall squared because rain also may
affect development in important ways.1
The four explanatory variables included in the regression models—
temperature, rainfall, education, and urbanization rates—explain a
very large part (73 percent to 92 percent) of the variation in income
levels among municipalities in Bolivia, Brazil, Chile, Mexico, and Peru
(table 9.1). This suggests that we have included the most important
explanatory variables, and that adding additional control variables would
likely make little difference.
Education level is the most important explanatory variable in all cases.
That is not surprising, given that the economic growth literature has
shown education to be a key growth-enhancing variable. The effect of the
two climate variables, temperature and rainfall, varies from country to
country as shown in figure 9.2.2
Panel (a) of figure 9.2 shows that average annual temperatures in
the Bolivian municipalities vary between 3°C and 27°C and that aver-
age annual consumption per capita ranges between US$245 and
US$2,562 (purchasing power parity, or PPP-adjusted US$ of 2001).
The relationship between temperature and consumption is nonlinear
and statistically significant, but the maximum is reached in the
warmest regions of the country, rather than in regions with temperate
climates. The coldest regions have consumption levels that are about
40 percent lower than those of the hottest regions, implying that the
cold parts of Bolivia might gain significantly if average temperatures
were to increase—all other things equal, and not taking into account
the issue of water scarcity discussed in chapter 2.
Panel (b) of figure 9.2 shows that the level of income in Brazilian
municipalities is significantly related to temperature, with average annual
Simulating the Effects of Climate Change on Poverty and Inequality 253
1515
30148
income per capita (PPA-$/year)
1315
25148
1115
20148
915
15148
715
515 10148
315 5148
115 148
5 7 9 11 13 15 17 19 8 10 12 14 16 18 20 22 24 26 28 30
average annual temperature (ºC) average annual temperature (ºC)
1087
887
687
487
287
87
8 10 12 14 16 18 20 22 24 26
average annual temperature (ºC)
seen that can be taken into account. In Brazil, for example, warming is
expected to be faster in the north than in the south—and that is what
has been observed over the last few decades (Timmins 2007). A central
projection would be a 2.5°C increase in temperatures in the northern
part of Brazil and a 2.0°C increase in the southern part during the next
50 years. Similarly, warming is expected to be faster in the northern part
of Chile than in southern Chile. In the following, the effects of 1.5°C of
warming in the northernmost part and 0.75°C in the southernmost part,
and intermediate values in intermediate regions of Chile, are analyzed.
In Peru, warming in the jungle region is expected to be around 2°C, in
the mountain region 1.5°C, and in the coastal region 1°C. In Bolivia,
warming is projected to be around 2°C in all lowland areas and 1.5°C
in highland areas. Finally, in Mexico the coast is projected to warm less
than the center, and especially less than the northern part of the central
region, but because no statistical relationship was found between climate
variables and income in that country, it makes no sense to simulate the
effect of changes in climate in Mexico.
With respect to precipitation, there is little agreement as to the direc-
tion of change, as the confidence intervals all include zero change (IPCC
2007b). The simulations of the effects of future climate change therefore
assume no change in precipitation, except in a region in central Chile
for which most models project a decrease in rainfall due to a poleward
shift of the South Pacific and South Atlantic subtropical anticyclones
(IPCC 2007a). The magnitude of decrease that can be expected in cen-
tral Chile over the next 50 years appears to be around 10 percent and
corresponds to what has been observed over the last 50 years (Andersen
and Verner 2008).
The estimated relationships from table 9.1 are used to simulate the
effects of the above-described changes in climate by comparing two sce-
narios: one with climate change (CC) and one with no climate change
(NCC). All variables other than climate variables are assumed to be the
same in the two scenarios, so as to isolate the effects of expected climate
change (see appendix B for the regression equations).
less (table 9.2). The north and northeast will lose 22.5 percent and
19.8 percent, respectively, according to this simulation.
Climate change can be expected to cause an increase in inequality
among Brazilian municipalities and to increase poverty. The largest sim-
ulated loss for any individual municipality is 29 percent for the munici-
pality of Uiramutã, in Roraima in the northern part of the Amazon. The
largest gain is 8.6 percent for the municipality of Campos de Jordão, in
the State of São Paulo. At the municipal level a strong positive relation-
ship appears (ρ = 0.58) between the current level of income and the
subsequent gains from climate change, indicating that currently richer
municipalities will likely lose less from future climate change than will
currently poorer municipalities. The simulations thus suggest that
future climate change is likely to contribute to an increase in inequal-
ity among Brazilian municipalities. Future climate change would also
act to increase poverty, as the regions that are currently the poorest are
all to see substantial income reductions due to temperature increases
(figure 9.3).
Figure 9.3 Brazil: Current per Capita Income versus Expected Future
Impacts of Climate Change, by Mesoregion
10
5
estimated effect of future climate
change on income (% change)
–5
–10
–15
–20
–25
–30
0 100 200 300 400 500 600 700
current per capita income (PPA-US$/month)
south south-east center-west north-east north
Figure 9.4 Peru: Estimated Change in Incomes Due to Future Climate Change
versus Current Incomes, Municipal Level
20
15
due to future climate change (%)
10
estimated change in income
–5
–10
–15
–20
–25
51 551 1,051 1,551
current income per capita (soles/year)
Table 9.4 Chile: Estimated Impact of Climate Change on per Capita Income,
2008–58
(percent)
Effect of temperature Effect of changes Total effect of future
Region changes in precipitation climate change
Tarapacá –8.3 0.0 –8.3
Antofagasta –7.9 0.0 –7.9
Atacama –7.4 0.0 –7.4
Coquimbo –7.1 0.0 –7.1
Valparaíso –6.8 0.0 –6.8
Metropolitan Region –6.8 0.0 –6.8
O’Higgins –6.6 0.0 –6.6
Maule –6.5 0.8 –5.7
Bío Bío –6.4 0.5 –5.9
Araucanía –6.2 0.7 –5.5
Los Lagos –5.9 –1.6 –7.5
Aisén –5.4 –0.6 –6.0
Magallanes –4.6 0.0 –4.6
Total –6.7 0.0 –6.7
Source: Authors’ estimations.
Simulating the Effects of Climate Change on Poverty and Inequality 261
Figure 9.5 Chile: Estimated Change in Incomes Due to Future Climate Change
versus Current Incomes, Municipal Level
0
change in incomes due to projected
–2
future climate change (%)
–4
–6
–8
–10
–12
–14
0 500 1000 1500 2000
current income per capita (U.S.$/month)
Source: Authors’ elaborations based on estimations.
Note: The size of the bubble indicates the population size in each municipality.
Summing Up
In this chapter municipality-level data from five large countries in the
LAC region were used to estimate contemporary relationships between
average climate and average incomes. The estimated relationships were
then used to simulate the impact of the climate changes that the IPCC
projects for the next 50 years. The simulation results should not be inter-
preted as forecasts but are simply indicative of the direction and magni-
tude of the effects that might be expected from the climate changes
suggested by the IPCC models. In the simulations, all other explanatory
factors than climate are held constant. That, of course, is not a realistic
assumption but simply serves to isolate the effect of climate change from
all the other factors that also will affect people’s incomes over the next
several decades. In reality, many things are going to change over the com-
ing decades, and most of the changes will likely reduce people’s vulnera-
bility to climate change. More educated people, for example, are usually
less vulnerable to climate change because their occupations are typically
less dependent on climate. Carbon dioxide fertilization, technological
advances, and expanded irrigation systems may work to make agriculture
less sensitive to climate change, too. That such changes are not taken into
account implies that the simulations carried out in this chapter may exag-
gerate the effects of climate change.
The results of the simulations are summarized in Table 9.6. Of the five
countries investigated, three are expected to be adversely affected (Brazil,
Chile, and Peru), as their average levels of income are estimated to fall,
Simulating the Effects of Climate Change on Poverty and Inequality 263
Notes
1. Notice that climate variability is not addressed in this model, only systematic
climate change over several decades.
2. All axes span the actual observed values of temperatures and income in each
country. More details about the data for each country can be found in the more
elaborate, individual country case studies (Andersen and Verner 2008, 2009,
2010; Andersen, Román, and Verner 2008; Andersen, Suxo, and Verner 2009).
References
Andersen, L. E., S. Román, and D. Verner. 2008. “Social Impacts of Climate
Change in Brazil: A Municipal-Level Analysis of the Effects of Recent Climate
Change on Life Expectancy, Consumption, Poverty, and Inequality.” Draft,
World Bank, Washington, DC.
Andersen, L. E., A. Suxo, and D. Verner. 2009. “Social Impacts of Climate Change
in Peru: A Municipal-Level Analysis of the Effects of Recent Climate Change
on Life Expectancy, Consumption, Poverty, and Inequality.” Policy Research
Working Paper No. 5091, World Bank, Washington, DC.
Andersen, L. E., and D. Verner. 2008. “Social Impacts of Climate Change in
Mexico: A Municipal-Level Analysis of the Effects of Recent Climate Change
on Life Expectancy, Consumption, Poverty, and Inequality.” Draft, World
Bank, Washington, DC.
———. 2009. “Social Impacts of Climate Change in Bolivia: A Municipal-Level
Analysis of the Effects of Recent Climate Change on Life Expectancy,
Consumption, Poverty, and Inequality.” Policy Research Working Paper No.
5092, World Bank, Washington DC.
———. 2010. “Social Impacts of Climate Change in Chile: A Municipal-Level
Analysis of the Effects of Recent Climate Change on Life Expectancy,
Consumption, Poverty, and Inequality.” Policy Research Working Paper No.
5170, World Bank, Washington, DC.
Dell, M., B. F. Jones, and B. A. Olken. 2008. Climate Change and Economic Growth:
Evidence from the Last Half Century. NBER Working Paper No. 14132.
Cambridge, MA: National Bureau of Economic Research.
Horowitz, J. K. 2006. “The Income-Temperature Relationship in a Cross-section
of Countries and Its Implications for Global Warming.” Department of
Simulating the Effects of Climate Change on Poverty and Inequality 265
both slow-onset hazards such as drought or sea level rise, and rapid onset
hazards such as hurricanes or floods. Other cases emphasize adaptation
measures that help people cope with their aftereffects.
The chapter focuses particularly on the part that social assets—namely
bonding, bridging, and linking social capital, as defined below—can play in
vulnerable communities and how they can affect a community’s ability to
engage with external entities such as nongovernmental organizations
(NGOs )and state institutions (local and federal) in times of distress.
As Adger (1999) notes, the way in which individuals and groups within
a society interact with one another will influence their vulnerability to
climate change, including variability and extremes, notably through
mechanisms such as risk sharing, mutual assistance, and collective action.
Social assets are often under stress just before, during, and after rapid-
onset climate-related disasters, but also in the case of slow-onset disasters
or changes, such as droughts or sea level rise, that cause continual stress.
As a result, social assets can be eroded, particularly when stress is pro-
longed or recurrent.
Thus far, attention to social assets in disaster management has been
rare, and the tendency has been to treat disaster management mostly as
an engineering issue calling for technical solutions.1 But different groups
are not exposed to natural hazards in equal measure. The vulnerability,
resilience, and livelihood assets of a community depend on its location-
specific biophysical, social, and economic circumstances. Vulnerability is
defined as the degree to which a system or community is susceptible to,
or unable to cope with or adapt to, adverse effects of climate change,
including climate variability and extremes (IPCC 2007). The opposite
side of the coin is resilience; the less vulnerable a community is, the more
resilient it is. Resilience is the ability of a system or community, be it
social or ecological, to withstand shocks and surprises and revitalize itself
when damaged (Tompkins and Adger 2004). In this chapter, the term
“coping capacity” refers to the short-term ability to safeguard oneself and
a given asset base, especially during and immediately following an
extreme climate-related event such as a flood or hurricane. “Adaptive
capacity” refers to the long-term planning and management of assets,
before or during recovery from such an event or in response to gradual
climate change. Clearly, vulnerability, resilience, and the ability to cope
vary from one community to another. This means that interventions to
help people prepare for, cope with, and adapt to the effects of climate
change need to focus not just on areas and numbers of people at risk, but
on who is at risk and the type of risk they face.
Building Short-Term Coping Capacity and Longer-Term Resilience 269
The chapter is structured as follows. The first of its three sections out-
lines the framework used for analysis. The SLF is extended to distinguish
among three different types of social capital that, in principle, communi-
ties can draw on when faced with natural hazards. Also, recognizing that
livelihood assets can be used either to avoid (or mitigate) the impact of
climate hazards on livelihoods or to adapt to their consequences, the dis-
cussion extends the SLF to accommodate a distinction between these two
different coping strategies: (a) enhancing the longer-term resilience of the
livelihood to a climate impact through adaptation, and (b) strengthening
coping and recovery capacity, as a more immediate-term response to a cli-
mate event, also through adaptation. In the second section of the chapter,
the augmented SLF is applied to five diverse cases: experiences with
droughts in southern Bolivia and in northeast Brazil illustrate the impor-
tance of increasing livelihood resilience through adaptation to slow-onset
climate-related disasters. Experiences with hurricanes in Nicaragua and
floods in Belize provide examples of successful livelihood adaptation
measures that improve short-term coping and recovery from sudden
extreme events. All five cases highlight the importance of social capital
and the role that institutions, especially local ones, can play in generating
synergy among the different types of capital and thus increasing the effec-
tiveness of coping strategies. Experiences reported in other chapters of
this book support these findings. The chapter concludes by drawing some
implications for policy, emphasizing the importance of approaches that
focus on the livelihood strategies of poor communities and also on local
governance structures, which play a key role in determining the impacts
of natural hazards on livelihoods.
Figure 10.1 Three Types of Social Capital for Communities Faced with Natural
Hazards
Social asset
Natural hazard adaptation and Natural hazard
resilience
Vulnerable Community
Bridging social capital: Linking social capital:
support and capital flows from support and capital flows from formal
external community and family government institutions (mitigation
members (remittances, migration policies, relief measures, development
opportunities) and NGOs (relief aid) projects, etc.)
They must also draw on their other scarce assets to cope. The fabric of
bonding capital tends to be strong in a vulnerable community, being
related to strong links of reciprocity or empathy, and so bonding capital
may be helpful for building specific reciprocity and mobilizing solidarity
within the community. But a risk exists that the closed-circuit relation-
ships on which it is based might be broken by a large climate shock.
Bridging capital has advantages for building links to external agents and
for disseminating information. However, although NGOs can assist in
creating links to public institutions and making claims for assistance, both
bonding and bridging capital are also important for effectively making
claims on public institutions. For instance, community mobilization can
be a key factor in obtaining the assets that strengthen resilience through
making claims on public officials and political actors. This applies both in
emergencies and in achieving access to important services on a routine
basis, helping build resilience.
Overinvestment in certain types of social capital could be counterpro-
ductive, and diversification among different types of social capital is of
great importance for both the individual agent and the larger society.
Inclusion and trust among some social groups could lead to (or show
themselves in) exclusion and mistrust toward other social groups. In addi-
tion, very strong bonding capital can generate islands of microsocieties
within the larger society (Adler and Kwon 2002; Putzel 1997). For an
individual agent, social capital is not costless to build, so it follows that
one can invest too much in maintaining social linkages and trust relations.
In analyzing susceptibility to adverse impacts of climate change, the
group of interest need not be an entire community determined geograph-
ically, but could be narrower, or could be defined on the basis of ethnic-
ity, religion, age, gender, occupation, and so forth. Hence it will often be
necessary to distinguish among the different types of social capital—and
in particular to focus on linking capital because vulnerable people are sus-
ceptible to discrimination and stigmatization from state institutions at all
levels.3 The following example illustrates the roles played by the different
types of social capital in a vulnerable community.
Such collective memory will often be crucial for preparing a given system
or location for change, building resilience, and coping with surprises.
For example, flooding is becoming increasingly common in poor urban
areas with limited social memory of disasters. As a result, as Moser and
Satterthwaite (2008) describe, responses tend to be mostly ad hoc, individ-
ualistic, and short term, such as sleeping on furniture, households’ digging
trenches around their property, and building small-scale water barriers.
Vulnerable groups can enhance their resilience if they can use linking
social capital routinely to make claims for access to basic services such as
sanitation, water supply, and flood defenses. Having rights to basic public
services and claiming those rights in ordinary times will strengthen the
general social resilience of a community and make it better prepared to
cope with adverse weather-based shocks. Citizenship is key in this process
because it provides ways to increase linking social capital. By the same
token, lack of citizenship (often the case for migrants) may leave people
with too little livelihood resilience against adverse events.
• Much physical capital was depleted by the disaster, and rebuilding was
still going on years later.
• Access to financial capital was limited after the disaster because of
lack of money, credit, and insurance.
• Human capital was severely affected or destroyed by both the direct
impact of the disaster in injuries and loss of life and the indirect effects
in the forms of disease and unemployment.
278 Rossing, Rubin, and Brisson
• The magnitude of the floods tore apart the social fabric of families and
whole communities; bridging social capital, in the form of remittances
from abroad, was an important source of income for the affected
households.
In situations where all community members face similar risks, the state
or local institutions can play an important role in (re)building social capi-
tal as well as physical infrastructure. These two aspects can be referred
to as “soft adaptation” and “hard adaptation,” respectively (Agrawal and
Perrin 2008). Strengthening linking social capital, where the state is
directly involved in the interrelation, is an important first step, as are
state efforts to help regenerate bonding social capital at the community
level. A logical next step would thus be to enhance the role of local
institutions in (a) addressing the potential underinvestment in social
capital (in particular bonding and bridging), so as to strengthen liveli-
hood resilience, and (b) regenerating social capital (in particular bond-
ing and linking) as part of an adaptation strategy.
In what follows, the augmented SLF framework is applied to four cases.
Two concern the need for enhancing longer-term livelihood resilience
through adaptation, and three address the need for enhancing immediate
or short-term coping capacity through adaptive measures. All emphasize
the importance of strengthening social capital and working with context-
specific measures at the local level.
2. Examining the extent to which the groups are capable of taking action
to avoid the effects of climatic impacts (predisaster damage limita-
tion), in anticipation, for example, of future hurricanes, or when faced
with a gradually worsening situation such as drought
3. Investigating the ability of different groups to cope with the immedi-
ate effects of natural hazards, that is, their ability to judge whether an
immediate response is required, especially when a natural hazard has
turned into an actual disaster
4. Analyzing how different groups are able to adapt over the long term
by undertaking planned actions after a slow- or rapid-onset climate
event (for example, rebuilding in a manner adapted to heightened ex-
posure or threat) in anticipation of similar or worse climatic impacts
in the future
disaster impacts on their economic, social, and natural capital. The initiative
was implemented from February 2006 to April 2007 to develop commu-
nity emergency preparedness with a risk management approach, targeting
5,500 community members and involving 500 people, including munici-
pal and local school officials. The following steps were used:
laborers in the region, little has been accomplished thus far in this regard.
Some resilience policies have sought to increase water supply, with some
success. However, as our use of the SLF forcefully shows, water supply is
just one of many factors determining livelihood resilience. In a context
still characterized by inequality of access to land and property rights, as
well as liquidity constraints, it is not enough to increase the natural
resource stock. Duarte and others (2006) concluded that government
interventions have consisted of postdrought social security and welfare
payments, with little investment in capacity building for community pre-
paredness. Lemos (2007) argues that instead of focusing mostly on
drought relief (coping), the government should also address longer-term
redistribution policies (adaptation that enhances resilience).
Together these experiences in southern Bolivia and northeast Brazil
illustrate the importance of adaptation to increase livelihood resilience
in communities confronted with climate-related hazards. From the south-
ern Bolivian case a key lesson is the importance of enhancing human
and different types of social capital, to ensure the longevity of project
progress. The situation in northeast Brazil emphasizes the importance of
focusing on building livelihood resilience through adaptation, as well as
enhancing coping capacity. Because many communities remain vulnera-
ble to climate change and variability, the current focus on postdrought
policies should be supplemented with investments in capacity building
for community preparedness.
during the emergency and the post-Mitch rehabilitation. Some mayors took
decisive steps to restore livelihoods but did so more as an outcome of their
natural leadership than based on any formally established arrangements
(Rocha and Christoplos 2001).
The importance of social memory in strengthening local capacity can
be seen in the disaster response in some of Nicaragua’s most conflict-
prone regions. For some municipalities, isolation brought on by armed
conflict during the 1980s had fostered and consolidated an organizational
capacity that the local population could use to tackle other kinds of prob-
lems successfully. The evacuation after Mitch could therefore be accom-
plished rapidly (Rocha and Christoplos 2001). Also, the mobilization of
provisions and organization of emergency shelters was very effective in
these municipalities, contradicting the common assumption that civil
society is weakest in war-torn areas. What seems to have happened is a
strengthening of bonding social capital, as a resilience strategy in response
to the depletion of other assets—most notably physical capital—already
caused by the armed conflict.
Despite difficulties, NGOs were able to re-create both bonding and
bridging social capital in certain sectors. Qualitative research by
Tomlinson (2006) in the village of El Mirador found deterioration of
social capital, but the same study also describes how NGOs have had
some success in generating social networks in health and education. The
study contrasts El Mirador’s experience with that of El Hatillo, another
village in the Matagalpa department that was more socially cohesive than
El Mirador. El Hatillo’s agricultural cooperatives and well-structured
women’s groups enabled the villagers to organize emergency committees
that brought in food and evacuated the sick for treatment in the wake of
Mitch. Seeing that they would receive little support from the municipal
or national government, El Hatillo villagers not only enhanced their
bonding social capital but also had various groups working constructively
to combine resources (bridging capital) to improve longer-term commu-
nity outcomes.
While cooperation between government and NGOs has managed
to increase the adaptive capacity of many poorer communities in
Nicaragua since 1998, a more sustainable social approach to climate
change must involve devolution of power from the central govern-
ment to municipalities. Closer collaboration between NGOs and the
municipalities could prove crucial in filling a current vacuum in the
Nicaraguan discourse on disaster mitigation and preparedness. Indeed,
NGOs generally can play key roles as intermediaries, creating links
288 Rossing, Rubin, and Brisson
The role of social capital in short-term coping with Tropical Storm Arthur
in Belize. The Stann Creek district in Belize was hit by Tropical Storm
Arthur in June 2008.16 The way in which the community, aided by local
government officials and NGOs, responded to this unprecedented flood-
ing disaster illustrates how different types of social capital, combined with
strong local governance, can be vital in people’s ability to cope with and
recover from a strong climate shock.
Arthur was unique in having no significant wind; low air pressure
caused the storm to hover over Belize for four days. The torrential rain
from the storm compounded the effects of Tropical Storm Alma, which
had developed in the eastern Pacific a few days earlier. Together, Arthur
and Alma brought more than 10 inches of rain across Belize within
36 hours, which led to the worst flooding the country has ever experi-
enced, affecting 80 percent of its population of some 300,000. As a result
of significant land clearing in the hills, insufficient land cover remained to
absorb the water. Moreover, trees and vegetation—natural protective
buffers—had increasingly been removed from riverbanks to allow better
vistas. Rivers in both southern and northern Belize overtopped their
banks causing flash floods, particularly in low-lying and coastal areas.
As Tropical Storm Arthur approached, people were expecting the
impact to come in the form of wind rather than water. Though people in
Belize are psychologically prepared for the storm and wind effects of hur-
ricanes, few had perceived that the country is actually much more vulner-
able to flooding, including flooding inland from heavy rainfall not
necessarily related to coastal storms, according to the country’s National
Emergency Management Organization (NEMO).
In Stann Creek, a poor rural district of about 10,000 people, most
of the villagers were caught by surprise. In previous floods rivers had
risen gradually and only to a few feet, but on this occasion water lev-
els rose several feet within a few seconds, trapping some people in their
houses and sweeping others away. The areas hardest hit were around
the poorest part of Stann Creek. More than 20 homes were destroyed
and 200–300 were severely damaged, displacing about 4,000 people.
However, only five people died.
That so few people died, and the prompt start to the postdisaster
process, can be directly attributed to high levels of social capital. Despite
the poverty of Stann Creek, all types of social capital were displayed
Building Short-Term Coping Capacity and Longer-Term Resilience 289
during the cleanup process. Bonding and bridging capital served as fragile
safety nets immediately following the disaster, and linking capital (ties
between the community and formal institutions) also played a critical
role. The local mayor was one of the first people on the spot. Possibly
because of the considerable trust that community members already
placed in him, he played a vital role in establishing shelters and organiz-
ing the distribution of emergency supplies, in close collaboration with the
army, which sent troops to the area almost immediately.
The bonding capital observed was particularly strong. In the immediate
aftermath, community members helped one another to the best of their
abilities. The least affected quickly and willingly reached out to people
who had lost everything, providing them with temporary shelter or taking
care of their children while the cleanup process was initiated. Belongings
and necessities such as food, water, and clothes were proactively
shared among the affected. The bridging capital with other communi-
ties and NGOs was also significant. The fact that fishermen from a
nearby village came almost immediately to the rescue in their boats
helped to save many lives.
Part of the reason why the recovery process began so promptly was that
members of the community, along with municipal officials and local
branches of various NGOs, knew how to organize quickly. The Red Cross
was already present in the village, undertaking training workshops, when
the disaster struck. A pilot program to train the community in emergency
preparedness had been started about six months before the disaster. This
program was part of a regionwide effort, spearheaded by the Red Cross, to
involve local communities in preventing and mitigating natural disasters.
The pilot program pioneered the creation of community networks, which
were put in charge of village risk and disaster management, coordinated
through the Dangriga branch of the Red Cross and the District Emergency
Management Organization. Even though the program had not yet
installed early warning systems, the necessary networks were already being
trained in the geographical mapping of hazards. They had also undertaken
an assessment of vulnerabilities differentiated by gender. Hence, when
Arthur struck, women participated actively in all relief operations. They
carried out rescue missions, rehabilitated local infrastructure (such as
schools), and along with men, distributed food.
In summary, these two cases emphasize how different types of social
capital, combined with strong local governance, can be vital in longer-
term resilience and short-term coping capacity when communities face
strong climate-related shocks.
290 Rossing, Rubin, and Brisson
Baseline data collected before the start of the program showed that
early cognitive and social development was severely delayed among chil-
dren aged 0–7 years, in very poor households. For example 97 percent of
those aged 3–7 years were in the lowest quartile with respect to language
development, and 85 percent were in the lowest decile, implying that
they were at least 21 months delayed in their receptive vocabulary. Large
delays were also found in the children’s short-term memory and in their
social development. Generally, the older children were the most delayed.
In surveys undertaken nine months after the program was initiated,
large improvements were found. School attendance rates for children aged
6–15 years were up. Nutrition and preventive health care were greatly
improved both in quality and quantity, and significant improvements
appeared in cognitive development outcomes of children aged 0–7 years,
particularly in language and social-personal skills. While improvements
were found among the beneficiaries of all three packages, compared to the
control group, which saw no improvements, the most pronounced effects
were found in the group that had received the nutrition and education
package together with productive investment grants.
In basic terms, all participant groups had their income protected
against natural disasters (such as drought, mudslides, and hurricanes),
while communities not in the program and hence not receiving any con-
ditional cash transfer would experience loss of income. This effect was
even more pronounced for consumption. The households in the group
that had received productive investment grants were best able to increase
their income and hence increase their savings, enabling them to cope bet-
ter with natural disasters. In other words, they were able to strengthen
their ex-ante risk management strategies. They were also less likely to
migrate than other groups, more likely to diversify their income sources
away from agriculture, and more likely to invest in improved technology.
A very interesting and positive effect was that people who received the
productive grant started planning for the future, whereas previously they
had lived a day-to-day existence. In the words of one respondent,
The evidence from such interviews suggests that aspirations and per-
spectives toward the future may be key for improving household welfare
and program impact. The interviews also indicate the potential role of
social interactions in changing attitudes. Because most of the households
in the communities covered—just a select few—received one of the three
packages, the community as a whole felt involved. Several information
meetings were held over the duration of the program, and a lot of discus-
sion occurred concerning what worked well and less well. Thus the pro-
gram, in addition to giving monetary assistance, also enhanced bonding
capital and fostered the building of bridging social capital. The outcome
of the education and nutrition income transfer was better livelihood
resilience (improved ex-post coping behavior), and the productive invest-
ment grant helped initiate livelihood adaptation by making people who
previously had focused on day-to-day survival more forward-looking.
Policy Perspectives
Although the causes of climate change are global, the adverse impacts
take the form of localized natural hazards, often at the subnational level.
Although measures to mitigate climate change and variability often
involve international negotiations and agreements and call for techno-
cratic, top-down solutions, measures to promote adaptation to the effects
of climate change to enhance longer-term resilience and short-term cop-
ing capacity are usually conceived at the national or subnational level.
Indeed they most often spring from the vulnerable people themselves,
either individually or as collective action (Agrawal 2008). The findings in
this book repeatedly emphasize the importance of actions with a local
focus. Though, as Agrawal (2008) notes, “information is currently lacking
on how and under which conditions area-based and decentralized devel-
opment approaches can help reduce climate change–related vulnerability,
enhance adaptive capacity, and promote sustainable livelihoods,” the
analysis in this chapter has shown that a fruitful way to inform this debate
from a social perspective is to use an augmented sustainable livelihoods
Building Short-Term Coping Capacity and Longer-Term Resilience 293
framework with a focus on social capital. The last section of this chapter
presents some implications for policy, drawing also on recurrent themes
from earlier chapters.
Notes
1. However, as noted by Nakagawa and Shaw (2004), the Kobe Earthquake in
1995 in Japan encouraged a new movement toward multidisciplinary emer-
gency responses, with clear links between technological and social solutions.
Consequently, in recent years, disaster management has become increasingly
connected to other sectors, particularly environment, city planning, and com-
munity participation. The change in approach stems from a growing realiza-
tion of the importance of incorporating people and communities in
predisaster mitigation and/or postdisaster recovery initiatives.
2. The classic study in this area is Conklin and Graham 1995.
3. Thanks to A. Norton for bringing our attention to this point.
298 Rossing, Rubin, and Brisson
4. The majority of social indicators for Chiapas are much lower than elsewhere
in the country. More than two-thirds of its total population is poor, compared
with 47 percent of the national population. Consequently, while the national
mean of population without school attendance in 2005 reached 8 percent, in
this state it exceeded 20 percent. The child mortality rate is 30 per 1,000 chil-
dren, compared to the national average of 22 per 1,000 children, and is
around 1.5 times the rate in Belize. Also, the state reported the highest agri-
cultural losses in Mexico due to climatic events over the past three decades
(Conde Saldaña, and Magaña 2007). The 2005 hurricane season was particu-
larly destructive, causing more than US$400 million in losses in the southeast
of Mexico, including Chiapas (Saldaña-Zorilla 2007).
5. See also the Nicaragua case study later in this chapter, which found that social
income transfers in the form of productive investment grants provided an
opportunity to start accumulating assets and hence improve the coping and
adaptive capacity of subsistence farmers.
6. Off-farm jobs and emigration are often the coping and adaptation strate-
gies (Conde, Saldaña, and Magaña 2007). In addition, nearly all farmers
receive a direct payment per hectare planted through government pro-
grams (PROCAMPO); however, this program was not designed as a rural
finance program (Wehbe and others 2006).
7. As argued in chapter 3 and supported by Stern (2007); Mendelsohn, Dinar,
and Williams (2006); and Heltberg, Jorgensen, and Siegel (2008), the impacts
of climate change fall disproportionately on poor and vulnerable people
who have contributed the least to cause the problem and who have the
fewest resources to cope with the impacts. While this issue is increasingly
considered one of social justice, it is beyond the scope of this chapter.
However, an interesting discussion of the social justice and human rights
aspects of climate change can be found in Mary Robinson’s Barbara Ward
Lecture (Robinson 2006).
8. Chapter 3 elaborated how social assets are often under stress just before, dur-
ing, and after climate-related disasters and how these assets can be eroded as
a result.
9. The poor tend to recover much more slowly than the nonpoor following an
exogenous shock, perpetuating their poverty and increasing inequality
(Baez and Mason 2008).
10. A step-by-step method for carrying out this identification is illustrated in
chapter 3.
11. Heltberg, Jorgensen, and Siegel (2008) provide a more detailed discussion of
social policy measures, such as “no-regrets” social policy and social protection
for adaptation, social funds for community-based adaptation, and social safety
nets for coping with natural disasters and climate shocks.
Building Short-Term Coping Capacity and Longer-Term Resilience 299
12. This case study is based on information from OCHA 2004; ISDR 2008; and
project information from CARE Bolivia’s Web site, http://www.carebolivia
.org/site_ingles/_pproyectos1.asp?id_pro=22.
13. Also see chapter 2.
14. This causes a substantial outflow of young people from the area.
15. CCER commissioned CIET (Centro de Investigación de Enfermedades
Tropicales) to conduct a three-phase social audit of the process, in an effort to
build the community’s voice into reconstruction efforts.
16. This case study is based on data obtained during a field visit to Belize in June
2008. Interviews were carried out with people affected by the disaster and
people assisting them in their recovery. Additional sources include Red
Cross/Red Crescent 2007; ReliefWeb 2008; and Woolcock 2000.
17. This case study draws on Baez and Mason 2008; Macours, Schady and Vakis
2008; Macours and Vakis 2008; and the World Bank Web site “Learning from
the ‘Atención a Crisis’ Pilot Program in Nicaragua’s Drought Region.” http://
web.worldbank.org/wbsite/external/countries/lacext/extlacregtoppovana/
0,,print:y~iscurl:y~contentmdk:21762443~pagepk:34004173~pipk:34003707
~thesitepk:841175,00.html.
18. For more info on process see http://unfccc.int/national_reports/napa/items/
2719.php.
19. This project was funded by the Trust Fund for Environmentally and Socially
Sustainable Development (TFESSD) under the auspices of the World Bank.
References
Adger, W. Neil. 1999. “Social Vulnerability to Climate Change and Extremes in
Coastal Vietnam.” World Development 27 (2): 249–69.
———. 2003. “Social Capital, Collective Action, and Adaptation to Climate
Change.” Economic Geography 79 (4): 387–404.
Adger, W. Neil, T. P. Hughes, C. Folke, S. R. Carpenter, and M. J. Rockstro. 2005.
“Social-Ecological Resilience to Coastal Disasters.” Science 309: 1036.
Adler, Paul S., and Seok-Woo Kwon. 2002. “Social Capital: Prospects for a New
Concept.” Academy of Management Review 27 (1): 7–40.
Agrawal, Arun. 2008. “The Role of Local Institutions in Adaptation to Climate
Change.” Paper presented at Workshop on Social Dimensions of Climate
Change, Social Development Department, World Bank, Washington, DC.
Agrawal, A., and N. Perrin. 2008. Climate Adaptation, Local Institutions, and Rural
Livelihoods. IFPRI Working Paper W08I-6. Washington, DC: International Food
Policy Research Institute.
300 Rossing, Rubin, and Brisson
Conclusion
Sanne Tikjøb and Dorte Verner
Climate change is the defining challenge of our time. More than an envi-
ronmental issue, climate change and variability threaten to reverse recent
progress in poverty reduction and economic growth. This book links carbon
dioxide (CO2) emissions and climate change to environmental degradation
and to implications for social and economic opportunities for development.
Environmental degradation affects water availability, land, fisheries, and
wildlife, with social implications affecting food security, livelihoods,
health, and habitat. Excessive stress from these factors may cause addi-
tional impacts such as conflict, migration, and increased poverty and
inequality. Hence, climate change is a threat to poverty reduction and if
not addressed will further exacerbate the vulnerability of the poor.
Social implications of climate change are already being felt in Latin
America and the Caribbean, and it is the poorest who are most affected.
Even if national and global efforts to address climate change improve, cur-
rent and future climate trends have considerable momentum, and they will
dramatically affect economic, human, and social development for years to
come. Poverty, inequality, water access, health, and migration are and will be
measurably affected by changes in the climate. The study reported in
this book found that many already poor regions are becoming poorer;
traditional livelihoods are being challenged; water scarcity is increasing,
305
306 Tikjøb and Verner
Building the asset base of the poor: Enhancing local livelihoods and access to public services
Objective Policy action Level of implementation
Actions to strengthen physical capital
Improve access to public services. Invest in public works to increase access to water, electricity, and sewerage Central and local government
connection.
Prevent erosion and landslides Build and promote the use of safe structures such as gabion baskets and Central and local government
and manage standing water. drainage channels. Re-vegetate unstable slopes, redirect settlers to safe
“invasion zones,” employ custodians to prevent settling and denuding of
unstable slopes.
Enhance governance and Implement integrated water resources management at the river-basin level. Central and local government and
management of water. Increase collaboration among water, climate, and development specialists. NGOs
Decentralize water management to local level and include all stakeholders in
dialogue.
Improve water access and use. Invest in canals, dams, and water-saving technologies and improve institutional Central and local government, civil
arrangements, including water rights, water user associations, water pricing. society
Improve access to mechanisms for purifying water and improve water-storage
practices.
Improve the protection of Incorporate climate considerations into tourism planning and development. Central and local government
infrastructure vital to the tourist Enhance building codes and implement policies that restrict development in
sector, such as coastal zones. coastal zones.
Undertake environmental impact assessments prior to tourism development.
Respect existing natural structures (for example, do not replace existing, natural
rocky beaches with white sand).
(continued)
307
Table 11.1 Summary of Policy Recommendations (continued)
308
Building the asset base of the poor: Enhancing local livelihoods and access to public services
Objective Policy action Level of implementation
Support agricultural livelihoods Implement policies that ensure long-term-sustainable solutions to improve Central government
and increase food security. agrochemicals, genetic diversity, energy use, and infrastructure.
Improve infrastructure, including food safety storage, crop storage, and
transportation.
Expand agricultural extension service.
Build lasting infrastructure: roads, Improve and enforce building codes and provide assistance to improve Central and local government and
housing, and buildings. building structures. NGOs
Build or upgrade public buildings, such as schools, to storm shelters and
hurricane-proof essential public buildings such as hospitals, police stations,
and government buildings.
Build infrastructure such as roads and other communications infrastructure and
invest in technology development.
Protect assets against natural Implement vulnerability analysis to identify and map vulnerable populations Central and local government
disasters. and areas where climate hazards occur.
Enhance early warning systems to reach remote populations.
Establish and strengthen early warning systems with the participation of
indigenous peoples and poor farmers; for example, set up a network of radio
transmitters in different communities to prepare for climate-related events in
coordination with community radio broadcasting.
Develop awareness campaigns and provide disaster risk reduction training to
communities.
Set up information services to facilitate access to resources and assets for
adaptation to hurricanes.
Enhance government policies that place legal requirements on national
institutions (including local governments) to prepare and update emergency
and disaster risk reduction plans.
Increase awareness of disaster risk through broadcasting programs.
Actions to strengthen human capital
Reduce health risks. Improve access to health services especially for the poor. Central and local government and
Reduce malnutrition by implementing or expanding nutritional programs and NGOs
education.
Increase knowledge of household hygiene and environmental factors.
Develop models for projecting disease patterns, and improve data collection.
Invest in or improve health surveillance systems.
Assess the relative burden of future climate change–related health impacts and
vulnerabilities at the national and subnational levels.
Increase the living standards of Increase access to safety nets, health care, and education, including Central and local government
the poor to help them adapt. environmental education.
Improve the quality of education and health care systems.
Raise local awareness of climate Educate communities about climate change to increase local capacity to read Central government, local
change and adaptation and interpret local climate change patterns and trends. government, universities, and
programs. Provide local access to relevant information and analysis. NGOs
Improve information systems and policies related to livelihood programs, social
safety nets, and indexed insurance.
Actions to strengthen social capital
Increase community participation Enhance different types of community social capital: Local government, NGOs, and
in local projects and policy Involve community in preparations for climate change and extreme events. communities
development. Decentralize decision making.
Include local communities in decision and implementation processes.
Increase community participation, voice coalition, and local governance.
Increase support for household and community initiatives.
Strengthen existing bonds such as community nodes and other informal institutions.
Create awareness campaigns on disaster risk reduction that will further
cooperation within communities.
309
(continued)
310
Building the asset base of the poor: Enhancing local livelihoods and access to public services
Objective Policy action Level of implementation
Create local, national, and Carry out vulnerability analysis at the local, national, and regional levels, taking Local government, NGOs, and
regional vulnerability maps, into account asset availability and distinctions between long-term trends and universities
including hazard-prone areas, natural hazards. The vulnerability analysis should apply a context-specific
and increase knowledge about framework to ensure a strong match between the needs of target groups and
vulnerable communities. planned interventions.
Increase social knowledge. Undertake poverty and social impact analysis to understand the social Central and local government,
implications of climate change, including violent and nonviolent conflict, NGOs, and communities
migration, inequality, and poverty across social groups (age, gender, ethnicity,
location).
Improve dialogue. Implement policies that promote synergistic relationships between vulnerable Local government
groups and institutions.
Improve relationship between Develop programs and capacity building to increase and improve the dialogue Local government, indigenous
local institutions and indigenous between indigenous communities and local governments. peoples
peoples.
Enhance state capacity to deal Implement crime-, violence-, and conflict-prevention programs. Central and local government,
with conflict. Develop plans for deploying military and police forces. NGOs, and communities
Actions to strengthen cultural capital
Implement policies that ensure Catalogue traditional knowledge and broaden access to the information. Local governments and NGOs
long-term-sustainable solutions
to improve and strengthen
cultural capital.
Improve livelihood resilience and Draw on local knowledge when devising adaptation measures. Central and local government
adaptation.
Enhance the use of traditional Mobilize the use of traditional indigenous peoples’ institutions (as distinct from Local government, NGOs, and
knowledge in indigenous indigenous formal, political organizations). community
communities. Develop strategies and concrete support oriented toward helping indigenous
communities adapt to climate change and variability that capture the diversity
of livelihood strategies and the role and efficiency of cultural institutions.
Development agencies and national meteorological services must begin with
respect for traditional knowledge and use it as a starting point for their projects.
Extend technology and agricultural advice that take local cultures into consideration.
Actions to strengthen natural capital
Improve livelihood resilience and Improve integrated natural resource management and tailor it to local Central and local government
adaptation. circumstances through formulation of location-specific policies and
institutional arrangements.
Increase water sources and Increase conservation of water supplies or implement water pricing in Central and local government,
access to them. urban areas. communities
Shift to less-water-intensive agriculture in rural areas.
Create new highland reservoirs.
Facilitate workshops to create local cooperation on water management.
Preserve the natural resource Strengthen agricultural practices to account for timing and location of crop Central and local government,
base. activities; improve crop protection practices to include crop rotation and NGOs, communities, individual
diversification of farm activities. farmers
Alter inputs such as crop varieties, fertilizer rates, or irrigation.
Use technologies that harvest and conserve water and soil moisture.
Manage water to prevent flooding, waterlogging, erosion, and nutrient leaching.
Use weather forecasting to reduce production risk.
Conserve entire hillsides and watersheds rather than individual plots.
Prevent deforestation in indigenous territories.
Improve land and soil quality. Create awareness and build sense of stewardship among farmers. Central and local government and
Raise risk awareness regarding the quality of land. NGOs
Build environmental resilience. Strengthen policies that protect natural resources. Central and local government and
311
Expand protected areas, such as marine protected areas or parks, using NGOs
integrated coastal zone management principles.
(continued)
Table 11.1 Summary of Policy Recommendations (continued)
312
Building the asset base of the poor: Enhancing local livelihoods and access to public services
Objective Policy action Level of implementation
Prevent climate impacts on crop Enhance research on the potential effects of increased migration, such as Universities and community
productivity in indigenous diseases affecting crops and decreased productivity.
communities.
Support agricultural livelihoods. Improve agricultural research capacity taking into account climate change. Central government
Implement long-term-sustainable solutions to improve quality and access to
land, soil, water, and nutrients and improve information systems.
Improve land and soil to enhance Promote alternative productive systems that can better cope with climate Local government, NGOs, and
agricultural productivity in change and variability. community
indigenous communities.
Actions to strengthen financial capital
Create alternative livelihood Establish marine protected areas and train and employ local populations as Local government, private sector,
opportunities. rangers, researchers, guides, and park managers. local communities, and NGOs
Create programs that offer training in alternative livelihoods to populations at
risk, such as fishermen.
Develop marketing plans that include activities such as cultural and community
tourism, as alternatives to natural resources–based tourism.
Increase access to credit and land titles for the local population.
Increase local agriculturists’ market access by applying small-scale fair trade
principles.
Ensure sustainable aquaculture by employing stringent environmental legislation.
Build safeguards against climate change into aquaculture, for example, by
building deeper ponds and identifying climate-resilient species.
Strengthen indigenous peoples’ Enhance government policies that place legal requirements on national Central and local government
management of natural institutions (including local governments) to increase access to land titles
resources. for indigenous peoples’ territories, so that the territories are left in the hands
of indigenous peoples to control and manage.
Conclusion 313
with implications for the amounts of water available for farming and live-
stock husbandry, domestic use, agriculture, power generation, and indus-
trial use. Warmer air temperatures affect the geographical range of disease
vectors such as malarial mosquitoes, with severe implications for human
health. They also affect the range and yields of crops, with implications
for the viability of traditionally grown crop varieties and for agricultural
practices, food production and trade, and food security. In adults, tempo-
rary malnutrition reduces body mass, immunity, and productivity, but
the results are rarely permanent. In children it can stunt growth,
impede brain development, or cause death. Hence, risk is increased that
climate change may cause an intergenerational downward spiral in
human potential. Higher air temperatures also cause human health
problems directly, including raising mortality rates among infants, the
elderly, and other vulnerable groups. Further, higher temperatures com-
bined with decreases in soil moisture lead to deforestation, adversely
affecting people’s livelihoods in multiple ways.
Rising sea-surface temperatures and higher levels of carbon in seawa-
ter affect the size of fish populations, the viability and migration patterns
of fish stocks, and the health of coral reefs and mangroves. Damaged
ecosystems affect the livelihoods of people who depend on their sustain-
able exploitation.
Increases in the intensity of natural hazards such as hurricanes, and
changes in their geographical distribution, lead to higher death tolls
and more damage to livelihoods, property, and production systems.
Weather-related disasters affect human health not only through
encouraging the transmission of diseases but also through higher inci-
dences of food insecurity, caused by the erosion of crucial environ-
mental and physical assets. Changes in the predictability of seasonal
weather patterns have big implications for both commercial and sub-
sistence agriculture, sometimes rendering traditional routines obsolete
and wiping out crops, with implications for food prices, nutrition, and
food security.
Changes in precipitation amounts and patterns lead to more
droughts—affecting rural livelihoods and food security—and more floods—
affecting livelihoods, property, production systems, and food security
for both rural and urban populations. Already the widening incidence
of drought has been a key reason why one-third of the region’s rural
young people have migrated to towns and cities over the past 20 years.
Both droughts and floods also augment the risk of water- and vector-
borne diseases.
Conclusion 315
Rising sea levels lead to more floods and storm surges, affecting liveli-
hoods, property, and production systems; more beach erosion, affecting
the habitability of coastal settlements; salinization of soil, limiting or elim-
inating its use for agriculture; and the intrusion of saltwater into aquifers
relied on for drinking water.
Clearly, the social impacts of climate change depend not just on biophys-
ical exposure but also on the vulnerability of people and institutions to
shocks. Vulnerability is shaped by the wider political economy of resource
use, and as would be expected, is particularly pronounced among the
region’s poor.
Below are reported the main dimensions of the social impact of cli-
mate change in Latin America and the Caribbean. For each dimension,
the starting point is the climatic changes outlined above. Each section
reviews the trends and dynamics and summarizes the key policy recom-
mendations that also appear in table 11.1.
Water Scarcity
Although abundant in the region as a whole, water is scarce at the local
level in the same areas. Increased glacier melt, reductions in rainfall, ris-
ing sea levels, and more frequent extreme weather events reduce both the
availability and quality of water for human use. Water scarcity in turn has
three major social implications: first, it affects domestic water use and
agricultural output, thereby increasing the risk of food insecurity; second,
it affects the range and transmission of vector- and waterborne diseases,
which adversely affects health; and third, it disrupts entire livelihood
models, which may lead to migration and conflict. The social impacts of
water scarcity are growing and are being felt at a local level.
By 2050, the number of people facing water scarcity in Latin
America and the Caribbean could rise from more than 20 million today
to more than 75 million. Some areas are particularly vulnerable to water
scarcity. Arid or semiarid rural subregions in northeast Brazil, Bolivia,
Argentina, and Chile will experience a 20 percent reduction in water
runoff due to reduced rainfall by 2050. In the Andean region, the melt-
ing of intertropical glaciers will severely restrict water availability, putting
close to 40 million people, or 70 percent of the Andean population, at risk
of losing water supply for drinking, farming, and energy generation by
2020. In urban centers with high concentrations of poverty, where fresh-
water availability is already low, population growth and dense urbanization
patterns, along with growing pressure for economic development, push up
the demand for existing water. In many downstream rural communities,
316 Tikjøb and Verner
Natural Disasters
The incidence of natural disasters is rising, and poor areas are hit the hard-
est. The incidence of major floods, droughts, and storms in Latin America
and the Caribbean has been rising, from roughly 100 in 1970–79 to more
than 400 in 2000–08. Patterns of actual and predicted natural hazards
show that disasters are more likely to develop in poor areas, even if the
hazard frequency there is lower. While weather-related hazards are more
frequent in South America, they cause disproportionately more fatalities
in Central America and particularly in the Caribbean.
Poverty makes people vulnerable to natural disasters, and natural dis-
asters make people vulnerable to poverty. More than 8.4 million people
in the LAC region live in the path of hurricanes, and roughly 29 million
live in low-elevation coastal zones, making them highly vulnerable to sea
level rise and saline intrusion into groundwater supplies, storm surges, and
coastal flooding. On one hand, the impacts of natural disasters are socially
Conclusion 317
differentiated, and the poor are most affected by fatalities and injuries
because they tend to live in areas with high risk of floods, landslides, or
droughts; because the quality of their housing is too poor to withstand
severe weather events; and because they lack resources to help them
quickly recover lost assets. The high density of urban slums makes the
urban poor more susceptible to disease outbreaks following natural disas-
ters, while the rural poor are vulnerable because of their high dependence
on natural resources. On the other hand, natural disasters erode the asset
base of poor households by destroying natural and physical assets, divert-
ing human capital, depleting financial resources, and straining social
assets, pushing the poor deeper into poverty. Thus, the impact of disasters
is superimposed on existing vulnerabilities and may compound the diffi-
culties faced by the poor.
Disaster adaptation policies should combine a mix of hard and soft
adaptation measures to strengthen public infrastructure and protect the
asset base of the poor. Supporting the infrastructure needs in essential
public service areas, such as schools, hospitals, and police buildings, as
well as safeguarding access to water, electricity, and sewerage connec-
tions, will help build resilience in local communities and thus prevent
natural hazards from turning into disasters. In addition, applying a strate-
gic focus to the infrastructure needs of the poorest, for example, by
building food and feed storage and safe livestock facilities, will help
protect their asset base during extreme weather events. In particular,
hazard risk management frameworks should focus on developing
social capital in the community by incorporating participation and
voice coalition in the design of natural disaster adaptation initiatives.
Key policy recommendations include the following:
Rural Livelihoods
The rural poor are particularly vulnerable to climate change because of
their high dependence on natural resources. Environmental deteriora-
tion of the natural resource base—for example through global warm-
ing, which affects crop yields and viability as well as fish migration
patterns—will directly affect families and communities that depend on
these food sources for nutrition as well as income generation. For the
rural poor, who often lack the human and financial capital to diversify
their livelihoods, the depletion of existing assets increases the risk of
poverty and internal migration. Particularly vulnerable are agrarian
communities and artisanal fishermen, as well as communities dependent
on ecotourism.
replace fossil fuels, affecting world food prices. Key policy recommenda-
tions to support climate change adaptation in agrarian livelihoods include
the following:
Urban Livelihoods
Their housing conditions make the urban poor vulnerable to extreme
weather events, especially hurricanes. Incoming migrants typically set-
tle in flimsy housing in densely packed shantytowns, on marginal
lands such as floodplains, unstable slopes, and low-lying coastal land,
which other people consider too risky for settlement. The location of
the 20 largest cities in the LAC region near coasts, slopes, swamps, and
other areas prone to flooding exposes them to natural hazards.
Hurricanes, torrential rainfall, floods, landslides, and other climate-related
events may cause severe damage to shantytowns or destroy them alto-
gether. The same disasters may also destroy vital roads and utilities,
cutting off poor neighborhoods from help. As discussed in the section
on health, after such disasters, lack of clean water and sanitation may
provoke disease outbreaks. Warmer mean temperatures also raise the
risk of disease.
The vulnerability of the urban poor is exacerbated by their lack of
assets and public services needed to cope with climate change. In the
megacities of Latin America and the Caribbean, few of the poor have
education beyond the primary level. Often lacking regular jobs, most
make their living in the informal sector. They typically lack adequate
access to public services such as piped water, sanitation, and electricity—
some because connection is not available, others because connection is
too expensive.
Although the urban poor do not directly depend on natural
resources for a living, their poor housing conditions, job insecurity, and
lack of human and other assets put them at risk. For example, large
shares of the urban poor use their dwellings for productive activities
and income generation. Today more than one in four people in the
region (27 percent), or 65 percent of all the region’s poor, live in urban
slums, and the number is expected to rise as the threat to rural liveli-
hoods forces more and more people to migrate to urban areas in search
of new ones.
Thus far, urban authorities in Latin America and the Caribbean have
been slow to adopt measures for adaptation to climate change. City author-
ities need to identify and map where climate hazards are greatest, target the
most vulnerable groups, and design adaptation measures accordingly. In the
short term, policies that focus on disaster preparedness will help build
local resilience to natural hazards. In the long term, policies that address
building the asset base of the poor through improved housing, education,
health, and infrastructure will develop the adaptive capacity needed to
322 Tikjøb and Verner
• Enhance human capital of the urban poor, notably the provision and
quality of education, health care, and other social protection services.
• Improve property rights of the poor.
• Upgrade high-risk housing and utilities infrastructure such as safe
water supply, sanitation, and electricity, and enforce proper building
codes.
• Update disaster preparedness plans to establish roles and responsibili-
ties, and develop plans of action for immediately before and after an
extreme weather event.
• Implement relocation programs to safe, alternative housing, with clean
water, sanitation, and electricity at affordable prices. Programs should
inform residents of the importance of vacating high-risk areas, while
being sensitive to the costs borne by the people affected, especially if
they have to move away from current income opportunities.
Human Health
Changes in temperature, precipitation patterns, and extreme weather
events have growing direct and indirect impacts on health. Climate
change is likely to expand the geographic range of vector-borne diseases
and lengthen their transmission season. Dengue fever could become one
of the major health risks resulting from climate change and variability,
and new, more virulent strains are emerging. This disease is already
extending its reach in Mexico and central South America, and by the
2050s its transmission rates are likely to have grown two to five times in
most parts of South America, putting many cities at risk, including São
Paulo, Caracas, and Mexico City. The risks from malaria are particularly
serious when it spreads into new areas where immunity levels are low.
Local changes in temperature, rainfall, and humidity are expected to
cause malaria to advance into areas not previously affected, while the dis-
ease is projected to become less common in some areas where it is cur-
rently endemic.
The expected increase in the incidence of floods will lead to more out-
breaks of waterborne diseases, and increasing temperatures will create con-
ditions allowing pathogens to multiply faster. Waterborne diarrheal
diseases, including cholera, thrive where the lack of safe drinking water and
sanitation makes good hygiene difficult. This situation is much exacer-
bated in the chaotic conditions following natural disasters. These are by far
Conclusion 323
and low adaptive capacity (for example, areas with poor land use or
underdeveloped health infrastructure).
Education seems to be a very effective, “no-regrets” policy to counter
adverse health effects from climate change. Regression analyses show that
the relationships between climate and health are outweighed by strong,
unequivocally positive relationships between education and health. In
Brazil, for example, the estimated adverse health effects of climate change
over the next 50 years could be countered by an increase in average edu-
cation level of just 1.1 year. Building adaptive capacity and reducing
health risks, especially among asset-deprived groups, are together an
important development issue facing the region today. Policies to reduce
climate-induced health risks should focus on improving capacity to mon-
itor and predict disease patterns, building the asset base of the poor, and
reducing the vulnerability of groups facing the greatest health risks:
• Develop models for projecting disease patterns, and research the rela-
tionships among health outcomes, climate variables, and socioeconomic
factors.
• Implement functioning health surveillance systems to monitor the
incidence and spread of diseases and guide the development and
evaluation of adaptation strategies.
• Combat water scarcity by preventing groundwater from being con-
taminated by the use of sewage or wastewater for irrigation or by
runoff during floods and heavy rain.
• Implement simple, community-based mechanisms for purifying and
storing water that prevent parasites from thriving.
• Implement low-cost programs to promote household hygiene, and
expand nutritional programs for poor households at risk of malnu-
trition.
• Improve access to health care for migrant populations, and improve
access to safe water and sanitation in urban shantytowns.
Migration
Internal migration in response to climate change is already occurring in
the LAC region and seems likely to increase. Across the region during
1988–2003, one-third of the rural population aged 15–29 migrated to
urban areas. The widening incidence of drought has been a key factor.
Currently most of the LAC region’s migrants are young men, but the region
now also has one of the world’s highest migration rates for women. Climate
change projections indicate that many communities across Latin America
Conclusion 325
Conflict
Climate change may exacerbate existing risks of conflict rooted in socioe-
conomic factors. That may happen for several reasons, for example, if cli-
mate change diminishes the supplies of food, water, forests, energy, and
land or encourages migration. Other dangers will occur if climate change
undermines the capacity of the state, for example, by raising the costs of
infrastructure in remote areas, thus limiting the reach of the state; by
eroding fiscal resources; or by weakening the legitimacy of a state if it can-
not provide basic needs for affected populations. If natural disasters are
more frequent and severe, chaotic conditions may provide opportunities
for rebel groups to challenge the government’s authority.
Whether violent conflict is induced by climate change will likely
depend on how scarcities are perceived and communicated by key polit-
ical actors. Though most LAC countries are relatively strong, middle-
income societies with accountable governments, the adverse impacts
of climate change and variability are likely to put further stress on
those already suffering from high income inequality and crime rates.
That could lead some factions within weak political systems to exploit
the situation rhetorically and mobilize groups for conflict over scarce
resources.
The political economy should be considered to the same degree as
environmental degradation as an underlying reason for violent conflict
due to climate change. Adapting effectively to climate change and
avoiding violent conflict will require both local and national govern-
ments to be inclusive of the most vulnerable groups in society, ensuring
that they are heard and represented in dialogues with other stakeholder
groups. Addressing underlying political tensions and socioeconomic
inequalities constitutes a “no-regrets” policy that will allow local com-
munities and individual households to build resilience against climate
change and conflicts that it might induce. Key policy recommendations
to prevent violent conflict from climate change include the following:
• Improve access for the poor to quality education, health care, and social
safety nets.
• Undertake poverty and social impact analysis to understand the social
implications of climate change, including violent and nonviolent con-
flict, migration, inequality, and poverty across social groups (age, gender,
ethnicity, location).
• Implement policies that promote dialogues between vulnerable
groups and local and national government institutions.
• Expand the agricultural research agenda to account for the impact of
climate change on agricultural output and productivity.
requires a greater role for national and global structures to support local
processes—thereby making climate change adaptation an integral part of
future development work.
Physical Capital
Recommended actions to strengthen physical capital focus on improving
public works and infrastructure, with the wider goal of creating access to
services for the most vulnerable people. They include preventing erosions
and landslides, particularly for urban settlers; decentralizing water man-
agement to the local level; protecting productive infrastructure such as
tourist areas and storage facilities for local harvest; separating sewer and
storm drain systems and increasing their capacity; and devising early
warning systems to save lives. Such efforts will have multiple effects and
will help protect against the spread of diseases and the loss of jobs and
other productive means, as well as increase food security and integration
with agricultural markets. An often overlooked dimension of adaptation
is the transfer of technology. Climate-smart communication technologies
334 Tikjøb and Verner
Human Capital
Improving the health and education of a population is the safest “no-
regrets” approach to enable long-term adaptation to climate change. As
seen above, studies from Brazil show that an extra two years of schooling
can mitigate the negative effects of climate change on income. As a short-
term coping strategy, the use of climate indicators in conditional cash
transfer programs can help ensure adequate nutrition levels in chronically
poor populations affected by climate change. Other programs to
strengthen human capital include raising awareness of climate change and
its associated impacts, as well as of the adaptation programs that are avail-
able to families.
Social Capital
Building and preserving social capital are ongoing processes. As an
intangible asset, social capital can be difficult to measure and docu-
ment. Yet investments in social capital should be a priority because they
can be a foundation for the allocation of other assets and can create
important positive spillover effects. Building social capital is a “no-regrets”
approach to ensure that an adaptation measure chosen for implementa-
tion is targeted, timely, and effective. Through participatory approaches
that facilitate the dialogue among local communities and the sharing of
local concerns with local and national institutions, social capital can
develop within and among stakeholder groups (bonding and linking).
And it can help to ensure that local concerns are considered in devising
the appropriate resources and strategies. Existing institutional frame-
works can assist in promoting this asset, for example, through poverty
and social impact analysis, which helps achieve a deeper understanding
of how local climate change affects inequality, poverty, and migration
among different socioeconomic groups. Establishing partnerships with
local stakeholders will help promote the sustainable implementation of
an adaptation project.
Cultural Capital
Drawing on cultural capital is critical to designing adaptation strate-
gies with sustainable outcomes. Cataloguing traditional knowledge,
broadening access to that information, and mobilizing its integration
with local adaptation plans is a win-win activity for both traditional
Conclusion 335
Natural Capital
Preserving the natural resource base must be part of both urban and rural
adaptation strategies. While this objective covers a broad and important
field, certain priorities are worth highlighting for their direct effect on
livelihood resilience: water, agriculture, and environmental resilience. First,
to conserve water as an asset requires action to implement water pricing,
switching to less-water-intensive agriculture, and creating new highland
reserves. Second, to sustain livelihoods that depend on agricultural produc-
tion, adaptation measures should seek to improve crop rotation and diver-
sification of farm activities. Farming techniques that promote adaptation
to new environmental circumstances should complement traditional
farming methods. Finally, promotion of sound environmental policies,
particularly for watersheds, marine protected areas, and coastal zones, is
needed to support the development of the environmental resilience on
which natural capital depends.
Financial Capital
Expanding access to financial services must be a priority in adaptation
programs. In Nicaragua, for example, a conditional cash transfer program
was introduced with a productive investment grant as part of a package
with basic nutrition and education. This enabled recipients to begin
adapting their livelihoods to the growing threat of drought. The innova-
tive part of the program consisted of adding a climate-risk dimension to
social protection programs that traditionally focus on the chronically
poor. The same principle can be applied to other adaptive measures, such
as making transfers contingent on the recipients’ dwelling in less-exposed
or less-vulnerable areas. Other noteworthy approaches to strengthening
financial capital include social funds and support for community-driven
adaptation, safety nets for coping with climate risks and natural disasters,
improving access to credit and land titles, and microinsurance and
indexed insurance. Finally, offering local people training geared toward
employment in climate-resilient jobs will help support the development
of much-needed alternative livelihoods. A successful program in Belize
trained local fishermen who could no longer make a living from fishing as
rangers, researchers, tourist guides, and park managers, with the added
benefit of building resilience for their entire communities.
336 Tikjøb and Verner
Note
1. For a deeper look at the impact of climate change on indigenous people and
on the use of traditional knowledge systems and institutions in adaptation
planning, see J. Kronik and D. Verner, Indigenous Peoples and Climate Change
in Latin America and the Caribbean (Washington, DC: World Bank, 2010).
APPENDIX A
339
340 Christensen
Box A.1
Climate Definitions
Often there is a considerable gap between common perceptions of what is meant
by “climate” and a more scientific, meteorologically based definition. That mismatch
gives rise to misleading statements and assessments of the role of humankind in
observed changes, even in high-level documents. To better guide the reader of the
present volume,the following provides a basic meteorological understanding of
what climate is.
Annual Global Mean Temperatures
0.6 14.6
0.4 14.4
difference (°C) from 1961–1990
0.0 14.0
–0.2 13.8
–0.4 13.6
–0.6 13.4
–0.8 13.2
1860 1880 1900 1920 1940 1960 1980 2000
period rate
annual mean years °C per decade
smoothed series 25 0.177±0.052
5–95% decadal error bars 50 0.128±0.026
100 0.074±0.018
150 0.045±0.012
(continued)
Climate Change and Climatic Variability in Latin America and the Caribbean 341
Climate is basically the statistical properties of weather over a long period. Its sim-
plest and best-understood elements are mean temperature and mean precipitation
amounts, whether on a monthly, seasonal, or annual basis. No law of physics dictates
exactly what time frame to use. That weather changes from one year to another and
possibly shows significant trends over some years does not fall within this definition of
climate. A long time period must be defined in such a way that comparisons between
two climate periods (two equally long periods) are close to invariant. That means that
year-to-year, or even decadal variations should show up only marginally. However,
even at longer time scales the statistics of weather do not seem to be entirely invari-
ant. In practice, climatologists in the first part of the 20th century decided to use a
period of 30 years as a compromise to balance the need for invariance in the condi-
tions from one period to another. This led to the definition of 30-year climate norms,
which started with the period 1901–30. The latest norm is for the period 1961–91. For
practical reasons, these periods are still used with rigor. Many climate variables, such
as annual temperature and precipitation, are compared with respect to the latest ref-
erence period. As climate turns out to vary significantly even between these so-called
norm periods, it also has become customary to look at long-term trends.
The figure in this box depicts a time-series of annual global mean temperatures
from 1850 to 2005. The overall increasing trend is portrayed in a number of ways
(see figure note). It is important to observe that the underlying data points show a
clear scatter around the overall trend. Trying to interpret the time evolution over too
short a period is made difficult and misleading by the considerable interannual
“noise.” For example, trying to make a linear trend line through the last 8–10 years
(from 1998 to the present) would result in a trend close to zero, or even a negative
trend, suggesting that the overall warming has stagnated. Such interim periods
have occurred previously. It is only with a longer time span that the real picture of
change clearly shows itself. Change need not be a constantly evolving phenome-
non; even a reverse signal at times need not contradict the long-term evolution.
The definition of “climate change and variability” is not just a question of
comparing climate periods; in general it must also involve regions large enough to
show a coherent picture of change. Only then is it possible to relate the climate trend
of the site to any large-scale change taking place in the region of concern. The geo-
graphical scale on which such coherence is found appears to be not much finer than
continental. Therefore, clear-cut and robust statements about observed climate
change and variability on a detailed country basis would be dubious if they are not
balanced with a view to the larger-scale trends both geographically and temporally.
Opposite trends at nearby stations may still accord with changes on a larger scale.
Source: Author.
342 Christensen
Latin America and the Caribbean are warming largely in line with the
global trend and are likely to continue doing so. Local exceptions to the
general tendency will appear, but available models do not give robust results
for every part of the region. In particular, many issues related to changes in
the Amazonas are still unresolved because important aspects of the interac-
tion between vegetation and climate are still little understood. Different
models also tend to behave differently in simulating the present climate
within the region, and that limits the ability to take the simulated responses
to anthropogenic forcings at their face value for the region. For this reason,
the best estimate of climate change and variability in the region comes from
assessing the results from many models, as was done by Christensen and
others (IPCC 2007). The present analysis tries to go behind some of the
statements provided there, by undertaking a comparative assessment of the
results of general circulation models (GCMs).
Precipitation
Data are less widely available for precipitation than for temperature, but
the information that is available shows that precipitation varies widely
within the LAC region. Figure A.2 depicts the geographical distribution of
change. The most noticeable large-scale, coherent pattern is a long-term
tendency toward drying in the tropics and subtropics (the Caribbean
and northern South America), while temperate (southern South America)
climates experience more precipitation.
For large parts of the region, data are not readily available for analysis. In
some countries efforts have been made to rescue and collect observational
data, though not necessarily to make them publicly available. Analysis of
these data confirms the broad tendencies, but trends may vary quite widely
even at the community level (scale of 1–200 km), with quite large positive
Climate Change and Climatic Variability in Latin America and the Caribbean 343
<-2.0 -1.7 -1.4 -1.1 -0.8 -0.5 -0.2 0 0.2 0.5 0.8 1.1 1.4 1.7 >2.0 °C per century <-1.3 -1.1 -0.9 -0.7 -0.5 -0.3 -0.1 0 0.1 0.3 0.5 0.7 0.9 1.1 >1.3 °C per century
IBRD 37787
MAY 2010
<-100 -100 -80 -60 -40 -20 -5 0 5 20 40 60 80 100 >100% per <-60 -60 -45 -30 -15 -3 0 3 15 30 45 60 >60% per
century century
IBRD 37788
MAY 2010
trends occurring next to negative ones. It is only when data are aggregated
over large regions that changes can be described in the context of global
warming (Zhang and others 2007).
Sea Level
Sea level in many regions of the world varies considerably on many time
scales. The shortest time variation and in some regions by far the largest sea
level changes are induced by tides. As these occur on a regular basis, how-
ever, natural and managed environments are adapted to this variability.
Globally, sea level has been rising for centuries, and during the 20th
century an average increase of 0.17 meters was measured (IPCC 2007).
Regionally, however, considerable variation appears (Figure A.3). Oceanic
circulation is complex; there is little reason to believe that the observed
geographical distribution of these changes can be directly scaled to cli-
mate change and variability scenarios because changing ocean conditions
(currents, salinity, and temperature) may cause less-obvious results along
Figure A.3 Geographic Distribution of Long-Term Linear Trends in Mean Sea Level
-0.4 0.0 0.4 0.8 1.2 1.6 2.0 2.4 2.8 3.2 3.6 mm/yr
IBRD 37789
MAY 2010
Recent Events
Most of Latin America and the Caribbean has experienced several
instances of severe weather in recent decades, damaging property, infra-
structure, and natural resources. Fatalities in connection with torrential
rain and hurricanes have been counted in thousands in this century alone.
The overall perception of these events is that they signal climate change
and variability. However, no formal scientific detection of these changes
at the regional level has been made. That makes it difficult to assess
whether or not those extreme events—whether individually or collec-
tively—are the results of general climate change and variability, as indi-
cated by the overall warming of the continent and adjacent areas.
People who have experienced severe and damaging weather events
generally wish to be able to assign them to some special factor. Yet
changes in climate or the recurrence of extreme events need not be
related to global warming. To formally attribute change or the occurrence
of particular events to a cause, a statistically sound number of events must
normally be considered. By nature, extreme events are rare at any given
location. Even events that occur over a large geographical region such as
Latin America and the Caribbean cannot simply be aggregated and stud-
ied as a whole because the chain of physical events leading to any one of
them is likely to differ from event to event and from location to location.
That precludes a simple statistical treatment of the data. Therefore,
recent reports of the IPCC have had very little to say about recent
changes at a regional level, and even less about the national or provincial
level, and their possible links with global climate change and variability.
Despite these limitations, people relate the impact of new events to
their memory of past occurrences. And as adverse changes create a need
for adaptation and mitigation measures, it is important to use past expe-
rience in preparing for potential hazards.
Against this backdrop, the array of models that yield robust projections
of the global rise in mean temperature yields much less robust estimates
of regional climate change and variability (IPCC 2007). To the extent that
assessments of future climate change and variability can be made, they are
the following:
• All of Central and South America is very likely to warm during this
century. The annual mean warming is likely to be similar to the global
mean warming in southern South America but larger than the global
mean warming in the rest of the area.
• It is uncertain how annual and seasonal mean rainfall will change over
northern South America, including the Amazon forest. In some re-
gions, qualitative consistency is seen among the simulations (rainfall
increasing in Ecuador and northern Peru, and decreasing at the north-
ern tip of the continent and in southern Northeast Brazil).
Data Distribution Center)2 largely derives from GCMs, which have lim-
ited ability to accurately simulate local-scale climates, especially with
regard to the key parameter of precipitation. GCM data are commonly
mapped as continuous fields, which do not convey the low skill of the
models for many regions, or are area aggregated, which renders the results
of little value for local application.
To help understand the potential accuracy of climate change and vari-
ability projections derived from climate models, it is imperative to compare
the performance of these models against the observed climate. Below are
summarized some of the main LAC climate characteristics, which a
model should be able to describe with some realism if its projections for
the future are to be credible.
South America
A complex variety of regional and remote factors contributes to the
climate of South America (Nogués-Paegle and others 2002). The tro-
pospheric upper levels are characterized by high pressure centered
over Bolivia and low pressure centered over northeast Brazil. At low
levels the Andes effectively block air exchanges with the Pacific
Ocean, but a continental-scale gyre transports moisture from the trop-
ical Atlantic Ocean to the Amazon region and then southward toward
extratropical South America. The South American low-level jet starts
a regional intensification of this flow, channeling it along the eastern
foothills of the Andes into the so-called Chaco low.4 The LLJ carries sig-
nificant moisture from the Amazonas toward southern South America,
and it is present throughout the year, but strongest during the austral
winter season (Berbery and Collini 2000; Vernekar, Kirtman, and
Fennessy 2003).
A clear warm season precipitation maximum, associated with the
South American Monsoon System (SAMS), dominates the mean sea-
sonal cycle of precipitation in tropical and subtropical latitudes. The
rainfall over northern South America is directly influenced by east-west
circulation patterns, and consequently tropical sea surface temperature
anomalies affect regions such as the Ecuador coast and north-northeast
Brazil.5 The SAMS is also modulated by incursions of drier and cooler
air from the midlatitudes over the interior of subtropical South
America (Garreaud 2000; Vera and Vigliarolo 2000). Rainfall anomalies
over subtropical South America are associated with regional feedback
processes and interactions among the topography, the SAMS, and the
midlatitude systems.
Another important feature, a regional part of the ITCZ, is the South
Atlantic Convergence Zone (SACZ)—a southeastward extension of cloudi-
ness and precipitation from southern Amazonas toward southeast Brazil
and the neighboring Atlantic Ocean. The SACZ reaches its easternmost
position during December, in association with high precipitation over much
of Brazil, a southeasterly flow over eastern Bolivia, and low precipitation
in the Altiplano. The variability of precipitation during the austral summer
352 Christensen
Europe
1.0 1.0
0.5
0.5 0.5
0.0
0.0 0.0
1900 1950 2000
1900 1950 2000 1900 1950 2000
Africa
Temperature anomaly (°C)
1.0
South America
Temperature anomaly (°C)
0.5 Australia
1.0 1.0
0.5 0.5
0.0 0.0
the near term, even though the near-term horizon is where the need to
know the evolution may seem most urgent.
The implication of these findings for interpreting locally observed
recent changes in temperature and other climate variables is that they are
subject to doubt. Observations may even suggest that local change is
opposite to the global or continental-scale behavior. This does not dis-
prove global or regional climate change and variability, but it does remind
us that even if global temperatures should increase by 3–4°C within this
century, many sites may not experience any of the change, whereas oth-
ers will see an even greater change.
Climate Change and Climatic Variability in Latin America and the Caribbean 355
Box A.2
Whereas climate change and variability projections for the distant future are
independent of the precise state of current climate, that is not the case for projec-
tions covering the next 1–2 decades. Currently, imperfect knowledge, primarily
about the state of the oceans, precludes prediction of seasonal weather patterns
much beyond 6 months, even on the very large scales.
Climate change and variability projections for the near future—despite being
independent of the emission scenario—generally become very uncertain as a re-
sult of the signal-to-noise issue described above and the limited knowledge that
researchers possess about the exact current state of the climate.
Robust statements about change on the regional scale are therefore only possi-
ble if the model projections also are physically sound, meaning that a certain effect
is expected because of large-scale changes in atmospheric circulation, moisture
content, or temperature change. These measures of change should be captured by
most, if not all, climate models. In most cases, it is therefore not possible to provide
formal, quantitative estimates of error for the projected values of change.
Source: Author.
processes. As they are generally too large to be used as a basis for convey-
ing quantitative information on regional climate change and variability, the
LAC region is further divided into Southern South America (SSA),
Amazon (AMZ), and Central America and Mexico (CAM).
This regionalization is very close to that initially devised by Giorgi and
Francesco (2000), but it includes additional oceanic regions and some
other minor modifications similar to those of Ruosteenoja and others
(2003). The objectives behind the original Giorgi and Francesco regions
were that they should have simple shape, be no smaller than the horizon-
tal wave length typically resolved by GCMs (judged to be a few thousand
kilometers), and where possible, should recognize distinct climatic
regimes. Although these objectives may be met with alternative regional
configurations, as yet no such options are well developed in the regional
climate change and variability literature.
Several common processes underlie climate change and variability in
a number of regions, and before discussing LAC countries individually,
it is relevant to summarize some of them. The first is a fundamental
consequence of warmer temperatures and the increase in water vapor
in the atmosphere. Water is continually transported horizontally by the
atmosphere, from regions of moisture divergence (particularly in the
subtropics) to regions of convergence. Even if the circulation does not
change, these transports will increase because of the increase in vapor,
and regions of convergence will get wetter and regions of divergence
drier. We see the consequences of this increased moisture transport in
plots of the global response of precipitation—where, on average, precip-
itation increases in the intertropical convergence zones, decreases in the
subtropics, and increases in subpolar and polar regions (Meehl and oth-
ers 2007). Regions of large uncertainty often lie near the boundaries
between robust moistening and drying regions, with different models
placing those boundaries differently.
Two other important themes in the extratropics are the poleward
expansion of the subtropical high-pressure systems responsible for the
overall long-term stable and dry weather conditions, and the poleward
displacement of the midlatitude westerly wind bands with the associ-
ated storm tracks. This atmospheric circulation response is often
referred to as the “excitation of the positive phase” of the Southern
Annular Mode. Superposition of the tendency toward subtropical dry-
ing and poleward expansion of the subtropical highs create especially
robust drying responses on the equatorward boundaries of the subtrop-
ical oceanic high centers.
358 Christensen
4
2.5 3.5
0° 3.0 2.5
3
2.5
2.5
2.5
3.0
20° 3.5 2
2.5
3.5 1.5
3.0
1.5 2.0 1.5 1
2.0
2.0 0.5
40°
2.0
0
2.0
1.5 2.0
-0.5
1.5
1.5 -1
1.5 2.0
60°
50%
20°
Precipitation Response (%)
30
20
15
0° 10
5
0
20° -5
-10
-15
40° -20
-30
-50
60°
20° 21 Models
Number of Models > 0
19–20
17–18
0°
14–16
20° 8–13
5–7
40° 3–4
1–2
60° 0
110° 90° 70° 50° 30° 110° 90° 70° 50° 30° 110° 90° 70° 50° 30°
IBRD 37792
MAY 2010
Temperature
The balance of evidence assessed by the IPCC leads to the following
statement:
All of Central and South America is very likely to warm during this
century. The annual mean warming is likely to be similar to the global
mean warming in southern South America but larger than the global
mean warming in the rest of the area (IPCC 2007).
This implies that temperatures in all seasons will continue to rise dur-
ing the 21st century. Some aspects of temperature-related events are also
expected to change. Unless the temperature increase is a result of entirely
new circulation patterns, changes in heat wave frequency and intensity
are expected. Likewise, there is an enhanced risk of changes in the sea-
sonality of some severe events because a higher temperature level in gen-
eral will tend to favor a longer warm season with possible related extreme
events. A clear example is the hurricane season (Box A.3).
Box A.3
Hurricanes
An important driver for tropical cyclones and hence hurricanes is the sea surface
temperature. A threshold of about 26°C determines whether a hurricane can be
formed. In a warmer world with everything else being equal, the hurricane sea-
son is likely to be prolonged and the area prone to hurricane development may
expand. However, a zone free of such systems will always remain near the equa-
tor, where atmospheric motions cannot support the development of intensive
storms. March 2004 saw the first hurricane ever identified in the southern Atlantic.
This hurricane severely affected the coastal zone of northeast Brazil. Such systems
could become more frequent because of higher sea surface temperatures in the
region.
In the tropics, particularly when the sun is close to its zenith, outgoing long-
wave radiative cooling from the surface to space is not effective in the optically
thick environment caused by the high water vapors over the oceans. Links to
higher latitudes are weakest at this time, and transport of energy by the atmos-
phere, such as occurs when the sun is less strong, is not an effective cooling
(continued)
360 Christensen
Precipitation
The IPCC’s Climate Change 2007: The Physical Science Basis (2007) con-
sidered only a few features of future precipitation change in Latin America
and the Caribbean to be robust. It stated:
Chile and Patagonia are influenced by the polar boundary of the subtrop-
ical anticyclone in the South Pacific, and they experience particularly
strong drying because of the combination of the poleward shift of circu-
lation and increase in moisture divergence. The strength and position of
the subtropical anticyclone in the South Atlantic is known to influence
the climate of southeastern South America. The projected increase in
rainfall in southeastern South America is related to a corresponding pole-
ward shift of the Atlantic storm track.
Some projected changes in precipitation (such as the drying over
east-central Amazonia and northeast Brazil and the wetter conditions
over southeastern South America) could be a partial consequence of the
El Niño–like response that is projected by many models (Meehl and
others 2007).6 That would directly affect tropical South America and
affect southern South America through extratropical teleconnections
(Mo and Nogués-Paegle 2001).
Extreme Events
Little research is available on extremes of temperature and precipita-
tion for the LAC region. Christensen and others (2007) estimated
how frequently the seasonal temperature and precipitation extremes
as simulated in 1980–99 are exceeded, using the A1B scenario from a
large ensemble of GCM simulations.7 They found that essentially all
seasons and regions will be extremely warm by the end of the century.
In Central America, the projected decrease in precipitation is accompa-
nied by more frequent dry extremes in all seasons. In Amazonia, models
project extremely wet seasons in about 27 percent of all summers
(December, January, February) and 18 percent of all autumns (March,
April, May) in the period 2080–99, with no significant change for the
rest of the year. For southern South America significant changes are
not projected in the frequency of extremely wet or dry seasons.
However, more careful analysis is required to determine how often
these wet and dry extremes are projected by individual models in the
large ensemble of GCM simulations producing these results before
making definitive conclusions about the likelihood of these changes in
extremes.
On the daily time scale, an ensemble of simulations from two atmos-
phere-ocean-coupled general circulation models was analyzed; both mod-
els simulate a temperature increase on the warmest night of the year that
is larger than the mean response over the Amazon basin but smaller than
Climate Change and Climatic Variability in Latin America and the Caribbean 363
the mean response over parts of southern South America (Hegerl and
others 2004). Concerning extreme precipitation, both models foresee a
stronger wettest day per year over large parts of southeastern South
America and central Amazonia and weaker precipitation extremes over
the coasts of northeastern Brazil.
Changes in extremes were analyzed based on multimodel simula-
tions from nine global coupled climate models (Tebaldi and others
2006). Figure A.6 depicts projected changes in extreme precipitation
and the number of consecutive dry days (Meehl and others 2007). The
general pattern of change suggests that nearly everywhere precipita-
tion intensity is increasing, enhancing the risks of flash floods. At the
same time, a clear tendency appears toward an increase in the number
of consecutive dry days almost everywhere. This suggests an increased
risk of droughts. These seemingly opposite tendencies are physically
consistent: in a warmer atmosphere, more moisture is available for pre-
cipitation in moist air than under present-day conditions. This will lead
to an enhanced possibility for more intensive precipitation events. On
the other hand, higher temperatures will also make the less-moist air
masses even drier, leading to fewer rainy days overall, and will thus
increase the chance of longer dry periods.
In Central America and in the Caribbean, a substantial contribution to
the intensive precipitation events is connected with tropical cyclone
activity. It is therefore relevant to keep in mind how the incidence of such
systems may change. Recent studies with improved global models, rang-
ing in resolution from about 100 to 20 km, suggest future changes in the
number and intensity of tropical cyclones (hurricanes). A synthesis of the
model results to date indicates, for a warmer future climate, increased peak
wind intensities and increased mean and peak precipitation intensities in
future tropical cyclones, with the possibility of fewer relatively weak hur-
ricanes and more numerous intense hurricanes. However, the total num-
ber of tropical cyclones globally is projected to decrease. The observed
increase in the proportion of very intense hurricanes since 1970 in some
regions is in the same direction but is much larger than predicted by
theoretical models.
Sea Level
Projected rises in the global average sea level at the end of the 21st
century (2090–99) relative to 1980–99 were estimated by the IPCC
for the six SRES marker scenarios. The results, given as 5 percent to
364 Christensen
4.0 B1
A1B
2.0
0.0 Precipitation
Intensity
-2.0
1880 1920 1960 2000 2040 2080 Year
-1.0 -0.5 0.0 0.5 1.0 std. dev.
B1
0.5 A1B
0.0
-0.5
Table A.1 Projected Global Average Surface Warming and Sea Level Rise
at the End of the 21st Century
Sea level rise
(meters, at 2090–99
relative to 1980–99)
Model-based range
Temperature change
(excluding future
(°C, at 2090–99 relative to 1980–99)a
rapid dynamical
Case Best estimate Likely range changes in ice flow)
Constant year 2000
concentrationsb 0.6 0.3 – 0.9 NA
B1 scenario 1.8 1.1 – 0.9 0.18 – 0.38
A1T scenario 2.4 1.4 – 3.8 0.20 – 0.45
B2 scenario 2.4 1.4 – 3.8 0.20 – 0.43
A1B scenario 2.8 1.7 – 4.4 0.21 – 0.48
A2 scenario 3.4 2.0 – 5.4 0.23 – 0.51
A1F1 scenario 4.0 2.4 – 6.4 0.26 – 0.59
Source: IPCC 2007.
a. These estimates are assessed from a hierarchy of models that include a simple climate model, several Earth
Models of Intermediate Complexity (EMICs), and a large number of Atmosphere-Ocean Global Circulation
Models (AOGCMs).
b. Year 2000 constant composition is derived from AOGCMs only.
21st century. Moreover, most attention was given to the A1B SRES marker
scenario.8 To give some qualitative information about likely changes in
time periods closer to the present, that is, the 2020s or the 2050s, it is nec-
essary to introduce scaling arguments, as a full utilization of the available
general circulation model information entering the IPCC assessment work
is not practically possible for the present book. As demonstrated in
Christensen and others (2007), using a scaling approach is reasonable, pro-
vided the climate parameters being addressed are robust and vary slowly
over time. This further implies that one should address decades rather than
a particular year. Thus, investigating 2020 should inform one about the
period 2010–30, and estimates for 2050 would refer to the period
2040–60; but even longer periods would be preferable (see box A.1).
Taking a scaling approach would imply that the local amplification
factor, compared to the global mean temperature change, is known.
Figure A.5, in combination with the global temperature increase for the
A1B scenario, provides the foundation for such a scaling argument.9 The
mean warming between the periods 1990–99 and 2090–99 can be esti-
mated to be 3°C. For the period representing 2020, this number equals
0.5°C, and for 2050, the number is about 1.5°C. These figures correspond
to scaling numbers of 0.17 and 0.5 respectively. This can be immediately
used to interpret the top row of temperature maps in figure A.5.
Precipitation is somewhat more subtle to deal with; as a gross measure,
one may still use the same scaling numbers, keeping in mind that the
boundaries between positive and negative change could well change
because of nonlinearity in the climate system and natural variability,
which tend to dominate as long as the temperature signal is weak.
Considering the projected change in extreme events, as characterized
by the analysis of consecutive dry days and precipitation intensity dis-
cussed in connection with figure A.6, some indications about the scaling
can be found from inspecting the time evolution of the globally averaged
change. Note that little or no change is found for the 2020s, whereas by
the 2050s the first clear signs of the climate change and variability signal
toward the end of the century could be expected, although perhaps only
showing through with about 50 percent of the power.
Further Research
To advance the scientific knowledge and awareness of past, contemporary,
and future climate for Latin America and the Caribbean, it is vital that
efforts are made to secure or rescue existing databases of climate records
Climate Change and Climatic Variability in Latin America and the Caribbean 367
in the region. All such data archives, no matter how small, are potentially
useful for this purpose. Databases should be made available to the inter-
national scientific community—at best free of charge, but if not, then
through collaborative international projects to analyze and exploit them.
It is also essential to encourage greater involvement by LAC region sci-
entists in climate analysis at the international level. At present, too many
efforts in climate analysis go only as far as the archives of national mete-
orological services, although a huge effort has been undertaken by the
World Meteorological Organization, and in particular the U.S. National
Oceanic and Atmospheric Administration/National Climatic Data
Center, to push LAC climatologists forward.10 Only by entering the inter-
national science arena will regional findings become useful in assisting
humankind to mitigate and adapt to anthropogenic climate change and
variability.
Regional climate change and variability projections for Latin America
and the Caribbean based on comprehensive analysis of climate only exist
in the form of output from cause-resolution GCMs, with a few excep-
tions. A strong demand exists for coordinated research exploring a range
of possibilities to provide information on regional or local climate change
and variability. That could be achieved by enabling concerted actions,
with participation from the LAC region as well as developed nations,
using the approach applied in such research programs as the North
American Regional Climate Change Assessment Program (NARCCAP)
(Mearns and others 2005) and the Prediction of Regional Scenarios and
Uncertainties for Defining European Climate Change Risks and Effects
(PRUDENCE), for example (Christensen and others 2007). These initia-
tives focus on skillful projections of regional climate change and variabil-
ity, providing not only estimates of changes in mean properties and their
variation, but also the scientific knowledge to permit a quantifiable
assessment of the uncertainties associated with the projections.
Notes
1. A1F1 describes a high-emission scenario with high economic growth, where
global population peaks around 9 billion in 2050 and declines to about
7 billion by 2100, with continued high greenhouse gas (GHG) emissions
resulting in cumulative emissions from 1990 until 2100 of 2,182 GtC. B1, in
contrast, describes a low-emission scenario, with a similar development in
population figures, also with high economic growth, but where the gains of
the economic growth to a large extent are invested in improved efficiency of
368 Christensen
References
Arguez, A., ed. 2007. “State of the Climate in 2006.” Bulletin of the American
Meteorological Society 88: S1–S135.
Bales, R. C., D. M. Liverman, and B. J. Morehouse. 2004. “Integrated Assessment
as a Step toward Reducing Climate Vulnerability in the Southwestern United
States.” Bulletin of the American Meteorological Society 85 (11): 1727.
Berbery, E. H., and E. A. Collini. 2000. “Springtime Precipitation and Water Vapor
Flux Convergence over Southeastern South America.” Monthly Weather
Review 128: 1328–46.
Christensen, J. H., T. R. Carter, M. Rummukainen, and G. Amanatidis. 2007.
“Evaluating the Performance and Utility of Regional Climate Models: The
PRUDENCE Project.” Climatic Change 81: supl. 1, 1–6, DOI:10.1007/
s10584-006-9211-6.
Dettinger, M. D., D. S. Battista, G. J. McCabe, R. D. Garreaud, and C. M. Bitz.
2001. “Interhemispheric Effects of Interannual and Decadal ENSO–like
Climate Variation on the Americas.” In Interhemispheric Climate Linkages,
ed. V. Markgraf. San Diego: Academic Press.
Doyle, M. E., and V. R. Barros. 2002. “Midsummer Low-Level Circulation and
Precipitation in Subtropical South America and Related Sea Surface
Temperature Anomalies in the South Atlantic.” Journal of Climate 15:
3394–3410.
Emanuel, K. 2003. “Tropical Cyclones.” Annual Review of Earth and Planetary
Science 31: 75–104.
Frich, P., L. V. Alexander, P. Della-Marta, B. Gleason, M. Haylock, A. M. G. Klein
Tank, and T. Peterson. 2002: “Observed Coherent Changes in Climate
Extremes during the Second Half of the Twentieth Century.” Climate
Research 19: 193–212.
Hegerl, G. C., F. W. Zwiers, P. A. Stott, and V. V. Kharin. 2004. “Detectability of
Anthropogenic Changes in Annual Temperature and Precipitation Extremes.”
Journal of Climate 17: 3683–700.
Garreaud, R. D. 2000. “A Multi-scale Analysis of the Summertime Precipitation
over the Central Andes.” Monthly Weather Review 127: 901–21.
Giorgi, F., and R. Francesco. 2000. “Evaluating Uncertainties in the Prediction of
Regional Climate Change.” Geophysical Research Letters 27: 1295–98.
Gray, W. M., 1984. “Atlantic Seasonal Hurricane Frequency. Part I: El Niño and
30 mb Quasi-biennial Oscillation Influences. “ Monthly Weather Review 112:
1649–68.
IPCC (Intergovernmental Panel on Climate Change). 2001. Special Report on
Emission Scenarios. Geneva: IPCC.
370 Christensen
———. 2007. Climate Change 2007: The Physical Science Basis. Contribution of
Working Group I to the Fourth Assessment Report of the IPCC. Geneva:
IPCC.
Kidson, J. W. 1988. “Interannual Variations in the Southern Hemisphere
Circulation.” Journal of Climate 1: 1177–98.
Levinson, D. H., ed. 2005. “State of the Climate in 2004.” Bulletin of the American
Meteorological Society 86: S1–S86.
Levinson, D. H., and J. H. Lawrimore, eds. 2008. “State of the Climate in 2007.”
Bulletin of the American Meteorological Society 89: S1–S179.
Liebmann, B., G. N. Kiladis, J. A. Marengo, T. Ambrizzi, and J. D. Glick. 1999.
“Submonthly Convective Variability over South America and the South
Atlantic Convergence Zone.” Journal of Climate 12: 1877–91.
Magaña, V., J. A. Amador, and S. Medina. 1999. “The Midsummer Drought over
Mexico and Central America.” Journal of Climate 12: 1577–88.
Mann, M. E., R. S. Bradley, and M. K. Hughes. 2000. “Long-Term Variability in the
El Niño Southern Oscillation and Associated Teleconnections.” In: El Niño and
the Southern Oscillation: Multiscale Variability and Its Impacts on Natural
Ecosystems and Society, ed. H. F. Diaz and V. Markgraf. 321–72. Cambridge,
U.K.: Cambridge University Press.
Mearns, L. O., R. W. Arritt, G. Boer, D. Caya, P. Duffy, F. Giorgi, W. J. Gutowski,
et al., 2005. “NARCCAP, North American Regional Climate Change
Assessment Program, A Multiple AOGCM and RCM Climate Scenario
Project over North America.” Preprints of the American Meteorological Society
16th Conference on Climate Variations and Change. January 9–13, 2005. Paper
J6.10, 235–38. Washington, DC: American Meteorological Society.
Meehl, G. A., T. F. Stocker, W. D. Collins, P. Friedlingstein, A. T. Gaye, J. M. Gregory,
A. Kitoh, et al. 2007. “Global Climate Projections.” In: Climate Change 2007:
The Physical Science Basis. Contribution of Working Group I to the Fourth
Assessment Report of the Intergovernmental Panel on Climate Change, ed.
S. Solomon, D. Qin, M. Manning, Z. Chen, M. Marquis, K. B. Avery, M. Tignor,
and H. L. Miller. Cambridge, U.K. and New York: Cambridge University Press.
Mo, K. C., and J. Nogués-Paegle. 2001. “The Pacific-South American Modes
and Their Downstream Effects.” International Journal of Climatology 21:
1211–29.
Nakicenovic, N., O. Davidson, G. Davis, A. Grübler, T. Kram, E. Lebre La Rovere,
B. Metz, et al. 2000. IPCC Special Report on Emissions Scenarios. Cambridge,
U.K.: Cambridge University Press.
Nogués-Paegle J., C. R. Mechoso, R. Fu, E. H. Berbery, W. C. Chao, T. C. Chen,
K. H. Cook, et al. 2002. “Progress in Pan American CLIVAR Research:
Understanding The South American Monsoon.” Meteorologica 27(1, 2): 1–30.
Climate Change and Climatic Variability in Latin America and the Caribbean 371
373
374 Christensen
change is only available in a very limited amount for this region (more
and more is emerging).
In the following, a condensed assessment of findings in observed
trends in temperature and precipitation climates, along with notions
of recent severe weather-related events, is provided for each of the
countries in Latin America and the Caribbean. The assessment largely
builds upon figures and tables available in the IPCC 4th Assessment
Report. Most of the material is included toward the end of that docu-
ment. In addition, information about recent events has been extracted
from the National Oceanic and Atmospheric Administration, National
Environmental Satellite, Data and Information Service, National Climatic
Data Center, and American Meteorological Society State of the Climate
reports 2001–2007 (Bulletins of the American Meteorological Society, or
BAMS 2005; 2006; 2007; 2008).
The countries are listed alphabetical by region, starting with Mexico, fol-
lowed by countries on the American isthmus, those in the Caribbean, and
finally countries in South America. If going country by country, the reader
may find some redundancies, which reflect the fact that for the smaller,
neighboring countries, the differences between their experienced as well as
projected climate changes are minor. Projected warming for each of the
countries is left to the reader to interpret based on the maps, which are
reproduced from the IPCC report in appendix A. It is assumed that issues
related to precipitation and extreme weather occurrences are of most
importance. For each country an ultra-short statement—an image—of the
most pressing climate issues of the future is provided.
Mexico
Recent trends. Overall temperatures have increased by 0.1–0.4°C per
decade within the last 25 years (between 0.5° and more than 2.0°C since
1901). Notably, some regions have seen negative temperature changes.
The geographical variation is large, and the long-term changes are most
advanced in the northwest near Baja California, where most warming has
occurred. Annual rainfall changes are well established throughout the
country but also with strongly irregular patterns of positive trends versus
negative trends. While the north shows trends of precipitation decreases,
the rest of the country has experienced a weak precipitation increase dur-
ing the last century. Sea level rise has reached 3 millimeters per year
(mm/year) toward the Gulf of Mexico, while the increase is less at the
Pacific coast.
Summaries of Likely Climate Change Impacts, by Country 375
SONORA MEXICO
Ojinaga To
f o
Chihuahua
Ca
Guaymas
lif
COAHUILA
or
BAJA
CALIFORNIA Loreto
SUR NUEVO Gulf of Mexico
Los Mochis 25°N
25°N Torreón Monterrey Matamoros
Saltíllo LEON
DURANGO
La Paz Culiacán TAMAULIPAS
SIN
Durango
ALO
Mazatlán
Cabo San Lucas Zacatecas SAN LUIS
OCEAN AGUASCALIENTES POTOSI Tampico
QUERÉTARO
NAYARIT Aguascalientes San Luis
Potosí YUCATAN Cancun
Tepic VERACRUZ Merida
GUANAJUATO Guanajuato HIDALGO
110°W Querétaro Cozumel
Puerto Vallarta Guadalajara DISTRITO FEDERAL
Pachuca TLAXCALA Campeche
20°N JALISCO MEXICO QUINTANA
MEXICO CITY Jalapa Bay of Campeche
M EXICO Colima
Morelia Toluca Tlaxcala
Puebla
Veracruz
ROO
Chetumal
COLIMA MICHOACAN Cuernavaca TABASCO CAMPECHE
Selected cities and towns PUEBLA Gulf of
Villahermosa
GUERRERO Honduras
State capitals Chilpancingo BELIZE
CHIAPAS
National capital 0 100 200 300 Kilometers
Acapulco
Oaxaca
OAXACA Tuxtla
Rivers Tehuantepec Gutierrez
0 50 100 150 200 Miles MORELOS Puerto Gulf of
Main roads Escondido Tehuantepec
GUATEMALA 15°N
Projection. See table 13.4 and figure 11.15 in IPCC Working Group II
(WGII; IPCC 2007b) for mean temperature and mean precipitation.
Mexico is located in a region dominated by atmospheric subsidence,
and the signals for precipitation are relatively robust across general cir-
culation models (GCMs) for the annual mean change, which seems very
likely to be strengthened in a warmer world. Seasonal signals are also
robust and point toward an overall reduction in June-July-August rainfall,
376 Christensen
Central America
Belize
Recent trends. Temperatures have increased by roughly from 0.1° to
0.2°C per decade over the last 25 years (about 0.5°C since 1901).
Rainfall shows no consistent trend and SLR has increased by about
1.5 to 2 mm/year.
85°
Chetumal
CENTRAL AMERICA
Belize Principal highways
Belmopan Selected cities
Flores
MEXICO BELIZE Gulf of National capitals
Honduras International boundaries
Puerto Iriona
Barrios 15°
La Ceiba
San Pedro Sula
Cobán
GUATEMALA HONDURAS 80°
Quetzaltenango Guatemala
Tegucigalpa Danli
Escuintla
Santa San
Ana Salvador
San
Caribbean
EL SALVADOR Miguel NICARAGUA Sea
León Rama
Managua
Lake Granada
Managua 10°
Lake
Nicaragua
PACIFIC OCEAN COSTA RICA Puerto Limón
Panama Colón
Puntarenas San Canal
José Panama
10°
P A N A M A
0 100 200 Kilometers Palmar Sur David
Gulf
COLOMBIA
Chitré
of
0 100 200 Miles Panama
90° 85° 80°
IBRD 37795
MAY 2010
Projections. See table 13.4 and figure 11.15 in WGII (IPCC 2007b) for
mean temperature and mean precipitation. Belize is located in a region
dominated by atmospheric subsidence, and the signals for precipitation
are relatively robust across GCMs for the annual mean change, which
seems very likely to be strengthened in a warmer world. It is likely that
during the next century the risk of hurricanes will increase, and more-
intense hurricanes will occur. Notably, it is not clear if there will be more
hurricanes per se. Sea level rise will compare with global increase.
Images: Somewhat drier conditions and increased risk of tropical storm
and hurricane incidents.
Costa Rica
Recent trends. Over the last 25 years, temperature increases have been
less than or about 0.1°C per decade (about 0.5°C since 1901). Rainfall,
in general, shows a positive trend, while SLR is about 1.5 to 2 mm/year,
with highest values occurring in the Atlantic coast.
Projections. See table 13.4 and figure 11.15 in WGII (IPCC 2007b) for
mean temperature and mean precipitation. The precipitation signal is not
very robust across the GCMs, which indicates that the country is near the
378 Christensen
El Salvador
Recent trends. Temperatures have increased less than or about 0.1°C per
decade over the last 25 years (about 0.5°C since 1901), and rainfall shows
a general positive trend. Yet reductions in rainfall are occurring inland. Sea
level rise has occurred at about 1.5 mm/year.
Projections. See table 13.4 and figure 11.15 in WGII (IPCC 2007b) for
mean temperature and mean precipitation. The signal is relatively robust
across GCMs. The upward trend in the frequency of intense precipitation
will continue but not at a strong and significant rate. The dry day fre-
quency trend will continue upward, and during the next century the risk
of intensifying hurricanes will increase. Notably, it is not clear if there will
be more hurricanes per se. Finally, SLR will compare with global increase.
Images: Somewhat drier conditions, and continuing risk of hurricane
incidents.
Guatemala
Recent trends. Temperatures have increased less than or about 0.1°C per
decade over the last 25 years (about 0.5°C since 1901). Rainfall shows no
consistent trend, and SLR is about 1.5 to 2 mm/year, with the highest
values on the Atlantic coast.
Summaries of Likely Climate Change Impacts, by Country 379
Projections. See table 13.4 and figure 11.15 in WGII (IPCC 2007b) for
mean temperature and mean precipitation. The signal is relatively robust
across GCMs. The upward trend in the frequency of intense precipitation
will continue but not at a strong and significant rate. The dry day fre-
quency trend will continue upward, and during the next century the risk
of intensifying hurricanes will increase. Notably, it is not clear if there will
be more hurricanes per se. Finally, SLR will compare with global increase.
Images: Somewhat drier conditions and continuing risk of hurricane
incidents.
Honduras
Recent trends. Temperatures have increased less than or about 0.1°C per
decade over the last 25 years (about 0.5°C since 1901). Rainfall in gen-
eral shows a positive trend, but inland areas in the eastern part experience
a reduction in rainfall. Sea level rise is about 1.5 to 2 mm/year, with the
highest values occurring on the Atlantic coast.
Projections. See table 13.4 and figure 11.15 in WGII (IPCC 2007b)
for mean temperature and mean precipitation. The signal is relatively
robust across GCMs. The upward trend in the frequency of intense
380 Christensen
precipitation will continue but not at a strong and significant rate. The
dry day frequency trend will continue upward, and during the next
century the risk of intensifying hurricanes will increase. Notably, it is
not clear if there will be more hurricanes per se. Finally, SLR will
compare with global increases.
Images: Somewhat drier conditions and continuing risk of hurri-
cane incidents.
Nicaragua
Recent trends. Temperatures have increased at about 0.1°C per decade
over the last 25 years (about 0.5°C since 1901). Rainfall in general shows
a positive trend that is not completely consistent. Sea level rise is about
1.5 to 2 mm/year, with highest values on the Atlantic coast.
Projections. See table 13.4 and figure 11.15 in WGII (IPCC 2007b) for
mean temperature and mean precipitation. The precipitation signal is
not very robust across the GCMs, which indicates that the country is
near the boundary between increase and decrease in precipitation. The
upward trend of precipitation frequency continues but not at a strong
and significant signal. The frequency of dry days will increase, and over
the next century the risk of intensifying hurricanes is likely to increase.
Notably, it is not clear if there will be more hurricanes per se. Finally,
SLR will compare with global increase.
Images: Possibly overall drier conditions and continuing risk of hurri-
cane incidents.
Panama
Recent trends. Temperatures have increased at about 0.1°C per decade
over the last 25 years (about 0.5°C since 1901). Rainfall in general
Summaries of Likely Climate Change Impacts, by Country 381
shows a positive trend, but it is not consistent across the country. Sea level
rise occurs at about 1 to 2 mm/year, with the highest values on the
Atlantic coast.
Projections. See table 13.4 and figure 11.15 in WGII (IPCC 2007b) for
mean temperature and mean precipitation. The precipitation signal is not
very robust across the GCMs, which indicates that the country is near the
boundary between increase and decrease in precipitation. The upward
trend of precipitation frequency continues but not at a strong and signif-
icant signal. The frequency of dry days will increase, and over the next
century the risk of intensifying hurricanes is likely to increase. Notably, it
is not clear if there will be more hurricanes per se. Finally, SLR will com-
pare with global increase.
Images: Possibly overall drier conditions and continuing risk of hurri-
cane incidents.
The Caribbean
Antigua and Barbuda
Recent trends. Temperatures have increased by roughly 0.1° to 0.2°C per
decade over the last 25 years (about 1.0°C since 1901). Rainfall shows
mostly an upward trend, while SLR is 2 to 3 mm/year.
80° 70°
UNITED
STATES CARIBBEAN
National capitals
Nassau
THE BAHAMAS International boundaries
0 100 200 Kilometers
Havana
0 100 200 Miles
CUBA 60°
20
20°
U.S. Virgin
Cayman Is. HAITI DOMINICAN Puerto Rico
Islands (US) Netherlands Antilles (Neth)
(UK) Port-au-Prince REPUBLIC
(US) ANTIGUA AND
Kingston Santo
JAMAICA Domingo Basseterre BARBUDA
ST. KITTS AND NEVIS St. John's
Guadeloupe (Fr)
DOMINICA
HONDURAS Caribbean Roseau
Martinique (Fr)
Sea ST. LUCIA Castries
Netherlands Antilles (Neth) Kingstown BARBADOS
NICARAGUA ST. VINCENT AND THE GRENADINES Bridgetown
Aruba (Neth)
St. George's GRENADA
Port-of-Spain
Caracas TRINIDAD AND
COSTA RICA R.B. de TOBAGO 10°
San José 80° COLOMBIA VENEZUELA 60°
IBRD 37796
MAY 2010
Projections. See table 13.4 and figure 11.15 in WGII (IPCC 2007b) for
mean temperature and mean precipitation. The signal is relatively robust
across GCMs. The trend for precipitation frequency is not clear. Yet the
intensity of downpours may increase in connection with hurricanes. Dry
day frequency trends are expected to continue upward. Over the next
century, an increasing risk of more of the most intense hurricanes is
expected. Notably, it is not clear if there will be more hurricanes per se.
Sea level rise will compare with global increase.
Images: Somewhat drier conditions, and increased risk of tropical
storm and hurricane incidents.
Barbados
Recent trends. Temperatures have increased roughly 0.1° to 0.2°C per
decade over the last 25 years (about 1.0°C since 1901). Rainfall shows
mostly an upward trend, while SLR is 2 to 3 mm/year.
Projections. See table 13.4 and figure 11.15 in WGII (IPCC 2007b) for
mean temperature and mean precipitation. The signal is relatively robust
across GCMs. The trend for precipitation frequency is not clear. Yet in
connection with hurricanes, intensity of downpours may increase. Dry
day frequency trends are expected to continue upward. Over the next
century, an increasing risk of more of the most intense hurricanes is
expected. Notably, it is not clear if more hurricanes per se will appear. Sea
level rise will compare with global increase.
Images: Somewhat drier conditions and increased risk of tropical storm
and hurricane incidents.
Cuba
Recent trends. Temperatures have increased by roughly 0.1° to 0.2°C per
decade over the last 25 years (about 1.0°C since 1901). Rainfall shows
mostly an upward trend, while SLR is 2 to 3 mm/year.
Projections. See table 13.4 and figure 11.15 in WGII (IPCC 2007b) for
mean temperature and mean precipitation. The signal is relatively robust
across GCMs. The trend for precipitation frequency is not clear. In con-
nection with hurricanes, intensity of downpours may increase. Dry day
frequency trends are expected to continue upward. Over the next cen-
tury, an increasing risk for more of the most intense hurricanes is
expected. Notably, it is not clear if there will be more hurricanes per se.
Sea level rise will compare with global increase.
Images: Somewhat drier conditions and increased risk of tropical storm
and hurricane incidents.
384 Christensen
Dominica
Recent trends. Temperatures have increased by roughly 0.1° to 0.2°C per
decade over the last 25 years (about 1.0°C since 1901). Rainfall shows
mostly an upward trend, while SLR is 2 to 3 mm/year.
Projections. See table 13.4 and figure 11.15 in WGII (IPCC 2007b) for
mean temperature and mean precipitation. The signal is relatively robust
across GCMs. The trend for precipitation frequency is not clear. In con-
nection with hurricanes, intensity of downpours may increase. Dry day
frequency trends are expected to continue upward. Over the next cen-
tury, an increasing risk of more of the most intense hurricanes is expected.
Notably, it is not clear if there will be more hurricanes per se. Sea level
rise will compare with global increase.
Images: Somewhat drier conditions and increased risk of tropical storm
and hurricane incidents.
Dominican Republic
Recent trends. Temperatures have increased by roughly 0.1° to 0.2°C per
decade over the last 25 years (about 1.0°C since 1901). Rainfall shows
mostly an upward trend, while SLR is 2 to 3 mm/year.
Projections. See table 13.4 and figure 11.15 in WGII (IPCC 2007b) for
mean temperature and mean precipitation. The signal is relatively
robust across GCMs. The trend for precipitation frequency is not clear.
Summaries of Likely Climate Change Impacts, by Country 385
Grenada
Recent trends. Temperatures have increased by roughly 0.1° to 0.2°C per
decade over the last 25 years (about 1.0°C since 1901). Rainfall shows
mostly an upward trend, while SLR is 2 to 3 mm/year.
Projections. See table 13.4 and figure 11.15 in WGII (IPCC 2007b) for
mean temperature and mean precipitation. The signal is relatively robust
across GCMs. The trend for precipitation frequency is not clear. In con-
nection with hurricanes, intensity of downpours may increase. Dry day
frequency trends are expected to continue upward. Over the next cen-
tury, an increasing risk of more of the most intense hurricanes is expected.
Notably, it is not clear if there will be more hurricanes per se. Sea level
rise will compare with global increase.
Images: Somewhat drier conditions and increased risk of tropical storm
and hurricane incidents.
Haiti
Recent trends. Temperatures have increased by roughly 0.1° to 0.2°C per
decade over the last 25 years (about 1.0°C since 1901). Rainfall shows
mostly an upward trend, while SLR is 2 to 3 mm/year.
Projections. See table 13.4 and figure 11.15 in WGII (IPCC 2007b) for
mean temperature and mean precipitation. Signal is relatively robust
across GCMs. The trend for precipitation frequency is not clear. But con-
nection with hurricanes, intensity of downpours may increase. Dry day
frequency trends are expected to continue upward. Over the next cen-
tury, an increasing risk of more of the most intense hurricanes is expected.
Notably, it is not clear if there will be more hurricanes per se. Sea level
rise will compare with global increase.
Images: Somewhat drier conditions and increased risk of tropical storm
and hurricane incidents.
Jamaica
Recent trends. Temperatures have increased by roughly 0.1° to 0.2°C per
decade over the last 25 years (about 1.0°C since 1901). Rainfall shows
mostly an upward trend, while SLR is 2 to 3 mm/year.
Projections. See table 13.4 and figure 11.15 in WGII (IPCC 2007b) for
mean temperature and mean precipitation. The signal is relatively robust
across GCMs. The trend for precipitation frequency is not clear. In con-
nection with hurricanes, intensity of downpours may increase. Dry day
frequency trends are expected to continue upward. Over the next cen-
tury, an increasing risk of more of the most intense hurricanes is expected.
Notably, it is not clear if there will be more hurricanes per se. Sea level
rise will compare with global increase.
Images: Somewhat drier conditions and increased risk of tropical storm
and hurricane incidents.
Summaries of Likely Climate Change Impacts, by Country 387
Projections. See table 13.4 and figure 11.15 in WGII (IPCC 2007b) for
mean temperature and mean precipitation. The signal is relatively robust
across GCMs. The trend for precipitation frequency is not clear. In con-
nection with hurricanes, intensity level of downpours may increase. Dry
day frequency trends are expected to continue upward. Over the next
century, an increasing risk of more of the most intense hurricanes is
expected. Notably, it is not clear if there will be more hurricanes per se.
Sea level rise will compare with global increase.
Images: Somewhat drier conditions and increased risk of tropical storm
and hurricane incidents.
St. Lucia
Recent trends. Temperatures have increased by roughly 0.1° to 0.2°C per
decade over the last 25 years (about 1.0°C since 1901). Rainfall shows
mostly an upward trend, while SLR is 2 to 3 mm/year.
Projections. See table 13.4 and figure 11.15 in WGII (IPCC 2007b) for
mean temperature and mean precipitation. The signal is relatively robust
388 Christensen
across GCMs. The trend for precipitation frequency is not clear. In con-
nection with hurricanes, intensity of downpours may increase. Dry day
frequency trends are expected to continue upward. Over the next cen-
tury, an increasing risk of more of the most intense hurricanes is expected.
Notably, it is not clear if there will be more hurricanes per se. Sea level
rise will compare with global increase.
Images: Somewhat drier conditions and increased risk of tropical storm
and hurricane incidents.
Projections. See table 13.4 and figure 11.15 in WGII (IPCC 2007b) for
mean temperature and mean precipitation. The signal is relatively robust
across GCMs. The trend for precipitation frequency is not clear. In con-
nection with hurricanes, intensity level of downpours may increase. Dry
day frequency trends are expected to continue upward. During the next
century, an increasing risk for more of the most intense hurricanes is
expected. Notably, it is not clear if there will be more hurricanes per se.
Sea level rise will compare with global increase.
Images: Somewhat drier conditions and increased risk of tropical storm
and hurricane incidents.
South America
Argentina
Recent trends. A temperature increase of 0.1°C per decade has been
observed north of Buenos Aires, but no change is seen to the south dur-
ing the past 25 years (0.5°C warming since 1901, except near Buenos
Aires, where temperatures have increased by roughly 1.5°C). Annual
Summaries of Likely Climate Change Impacts, by Country 389
Managua NICARAGUA
70° 60° 50° 40°
Caracas
COSTA RICA Panamá
R.B. de GUYANA ATLANTIC 10°
San José VENEZUELA Georgetown
PANAMA Paramaribo OCEAN
Bogotá French Guiana (Fr)
COLOMBIA SURINAME
Quito 0°
0°
ECUADOR
BRAZIL
PERU
10°
10°
Lima
BOLIVIA Brasília
La Paz
20°
20°
CHILE PARAGUAY
PACIFIC Asunción
OCEAN
SOUTH AMERICA
A DISPUTE CONCERNING SOVEREIGNTY
0 250 500 750 1000 Kilometers OVER THE ISLANDS EXISTS BETWEEN
ARGENTINA WHICH CLAIMS THIS
SOVEREIGNTY AND THE U.K. WHICH
ADMINISTERS THE ISLANDS.
0 200 400 600 Miles National capitals
International boundaries
100° 90° 80° 70° 60° 50°
IBRD 37797
MAY 2010
a positive trend in the frequency of very heavy rains. Glaciers are retreat-
ing drastically.
Projections. See table 13.4 and figure 11.15 in WGII (IPCC 2007b) for
mean temperature and mean precipitation changes. The precipitation
signal is quite robust across GCMs. While the central part is likely to
become drier, wetter or unchanging conditions may occur in the north
and in the south. The upward trend in precipitation frequency contin-
ues but not at a strong and significant signal. The dry day frequency
trend generally continues upward, except in the far south. Over the
next century, the risk for more of the most intense extratropical
cyclones and associated storm surges will increase. Sea level rise will
compare with global increase. However, greater sea level rise is esti-
mated to occur near the mouth of River Plata. Finally, glaciers will
decrease in size and volume.
Images: Wetter in the north, drier central, and increased risk of drought
and heavy rains.
Bolivia
Recent trends. Temperatures have increased at about 0.1° to 0.2°C per
decade over the last 25 years (around 0.5°C since 1901, but information
only sparsely available). Rainfall shows a varying trend, which is consis-
tent across the Andes. In particular, reductions in rainfall are seen toward
the west.
Projections. See table 13.4 and figure 11.15 in WGII (IPCC 2007b) for
mean temperature and mean precipitation changes. Note that the precip-
itation trend will be mainly positive and the upward trend in precipitation
intensity will continue. The dry day frequency trend is also expected to
continue upward.
Summaries of Likely Climate Change Impacts, by Country 391
Brazil
Recent trends. Temperatures have increased by 0.1° to 0.4°C per
decade within the last 25 years (between 0.5° and more than 2.0°C
since 1901). The geographical variation within the country is large, and
long-term changes are expected to occur mostly in the southeast,
where the greatest warming has occurred so far. Annual rainfall changes
are not well established throughout the country. In general, however,
rainfall has decreased in the northern part of Amazonia, whereas both
positive and negative changes are seen in southern Amazonia. Toward
the southeast, precipitation has been rising significantly. Sea level rise
has reached 4 mm/year in several ports, around 3 mm/year around the
river mouth of Amazonas, 2 to 3 mm/year on the southeast coast, and
around 2 mm/year elsewhere.
Projections. See table 13.4 and figure 11.15 in WGII (IPCC 2007b)
for mean temperature and mean precipitation. The precipitation sig-
nal is not very robust across GCMs for the annual mean change, indi-
cating that the region is near the boundary between increase and
decrease in precipitation. Seasonal signals are more robust and point
toward an overall reduction in June-July-August rainfall—possibly turn-
ing into an increase toward the Andes countries. The upward trend in pre-
cipitation frequency continues, possibly with the exception of areas along
the east coast. The dry day frequency trend continues upward, and SLR
will compare with global increase.
392 Christensen
Images: Possible overall drier conditions and increasing risk for drought
and heavy rains.
Chile
Recent trends. Modest temperature increases of 0.1°C per decade have
occurred in the north, while no change, or even a 0.1°C cooling per decade,
has occurred toward the south over the last 25 years (often less than 0.5°C
of either warming or cooling, so no significant trend since 1901). Annual
rainfall has generally decreased, by as much as by 50 percent in the central
part of Chile. Sea level rise is on the order of 1 to 2 mm/year, with higher
levels in the southern part.
Projections. See table 13.4 and figure 11.15 in WGII (IPCC 2007b) for
mean temperature and mean precipitation. The precipitation signal is quite
robust across GCMs, emphasizing drying everywhere except the southern-
most part. Overall the picture is constant throughout the year. The upward
trend for precipitation frequency does not appear significant—in fact a
reduction seems more likely. The dry day frequency trend continues
upward. Additionally, over the next century the risk for more of the most
intense cyclones will increase. Sea level rise will compare with global
increases.
Images: Overall drier conditions and enhanced risk for droughts.
Glaciers are disappearing or being reduced considerably in volume and
area coverage.
Colombia
Recent trends. Temperatures have increased by 0.1° to 0.2°C per decade
over the last 25 years (0.5° to 1.0°C since 1901, but information is only
sparsely available). Annual rainfall shows no clear trend throughout the
country, though a negative trend has generally been observed in northern
Amazonia. Sea level rise is roughly 2 to 3 mm/year toward the Atlantic
and around 1 mm/year in the Pacific.
Summaries of Likely Climate Change Impacts, by Country 393
Projections. See table 13.4 and figure 11.15 in WGII (IPCC 2007b) for
mean temperature and mean precipitation. Note that the precipitation
trend will be mainly positive in the Andes, while somewhat negative in
Amazonia. The signal is relatively robust across GCMs. The country is
located near the dividing line; hence the significance of the exact position
of the divide between wetter and drier conditions is speculative. The pre-
cipitation intensity trend and the dry day frequency trend continue
upward, but the trends are not robust everywhere. In fact, the Pacific part
of the country may realize a reduction in dry days. Sea level rise will com-
pare with global increases.
Images: Wetter conditions in the west but with large interannual vari-
ability further east; increased chance for fast floods; disappearing glaciers.
Ecuador
Recent trends. Temperatures have increased by 0.1°C per decade over the
last 25 years (between 0.5° and 1°C since 1901, but coverage is poor).
Annual rainfall has generally been increasing, and SLR has occurred at
about 1 mm/year.
Projections. See table 13.4 and figure 11.15 in WGII (IPCC 2007b) for
mean temperature and mean precipitation. The signal is relatively
robust across GCMs. A clear tendency for wetter conditions is showing
year-round. The upward trend in precipitation intensity continues,
394 Christensen
while the trend in dry day frequency is turning downward. Hence the
country may realize a reduction in dry days. Sea level rise will compare
with global increases.
Images: Wetter conditions, increased chance for fast floods, disappear-
ing glaciers.
French Guiana
Recent trends. Modest temperature increases of 0.1°C per decade have
been observed within the last 25 years (about 0.5°C since 1901).
Annual rainfall changes are not well established, and SLR occurs at
about 2 mm/year.
Projections. See table 13.4 and figure 11.15 in WGII (IPCC 2007b) for
mean temperature and mean precipitation. Note that the precipitation
trend will be mainly negative, but the signal is not very robust across
GCMs, particularly toward the south. The country is located near the
dividing line between wetter and drier conditions for some parts of the
year. Hence, the significance of the exact position of the divide between
wetter and drier conditions is speculative. Precipitation intensity will
show a downward trend, while the dry day frequency trend continues
upward. Sea level rise will compare with global increases.
Images: Drier conditions but with some interannual variability.
Guyana
Recent trends. Modest temperature increases have occurred at about 0.1°
to 0.2°C per decade in the last 25 years (about 0.5° to 1.0°C since 1901).
Annual rainfall changes are not well established, and sea level rise is
occurring at roughly 2 mm/year.
Projections. See table 13.4 and figure 11.15 in WGII (IPCC 2007b) for
mean temperature and mean precipitation. Note that the precipitation
trend will be mainly negative, but the signal is not very robust across
GCMs, particularly toward the south. The country is located near the
dividing line between wetter and drier conditions for some parts of the
year. Hence, the significance of the exact position of the divide between
wetter and drier conditions is speculative. Precipitation intensity will
show a downward trend, while the trend in dry day frequency continues
upward. Sea level rise will compare with global increases.
Images: Drier conditions but with some interannual variability.
Paraguay
Recent trends. Temperatures have increased by 0.2°C per decade over the
last 25 years (about 1°C since 1901), and annual rainfall has significantly
increased as well.
Projections. See table 13.4 and figure 11.15 in WGII (IPCC 2007b) for
mean temperature and mean precipitation. Note that the precipitation
trend is close to zero, although the signal is not very robust across GCMs.
That is expected, however, since small positive and negative values would
tend to dominate the disagreement. The country is located near several
dividing lines between wetter and drier conditions. Hence, the signifi-
cance of the exact position of the divide between wetter and drier condi-
tions is speculative. Precipitation intensity shows an upward trend, and
the trend of dry day frequency continues upward as well.
Images: Relatively unchanged precipitation regimes but with increas-
ing risk for extreme rain and drought.
396 Christensen
Peru
Recent trends. Temperatures have generally increased by 0.1°C per decade
over the last 25 years (between 0.5 and 1°C since 1901, but coverage
is poor). Annual rainfall has generally been decreasing, except toward
the northwest, where an increase is seen. Sea level rise is 1 to 2 mm/year,
with highest values in the southern part.
Projections. See table 13.4 and figure 11.15 in WGII (IPCC 2007b) for
mean temperature and mean precipitation. Note that the precipitation
trend will be mainly positive throughout the country, though with some
variation showing a negative trend in the southern part. The signal is rel-
atively robust across GCMs. The country is located near the dividing line;
hence the significance of the exact position of the divide between wetter
and drier conditions is speculative. The precipitation intensity trend, as
well as the trend of dry day frequency, continues upward but is not robust
everywhere. In fact the northern part of the country may realize a reduc-
tion in dry days. Sea level rise will compare with global increases.
Images: Generally wetter conditions but with large interannual vari-
ability farther south; increased risk of fast floods.
Suriname
Recent trends. A modest temperature increase of 0.1°C per decade in the
last 25 years (about 0.5°C since 1901) has been observed. Annual rainfall
changes are not well established, and sea level rise has occurred at about
2 mm/year.
Projections. See table 13.4 and figure 11.15 in WGII (IPCC 2007b) for
mean temperature and mean precipitation. Note that the precipitation
trend will be mainly negative, but the signal is not very robust across
GCMs, in particular toward the south. The country is located near the
dividing line between wetter and drier conditions for some parts of the
year. Hence the significance of the exact position of the divide between
wetter and drier conditions is speculative. Precipitation intensity will
show a downward trend, while the trend in dry day frequency continues
upward. Sea level rise will compare with global increases.
Images: Drier conditions but with some interannual variability.
Projections. See table 13.4 and figure 11.15 in WGII (IPCC 2007b) for
mean temperature and mean precipitation. Note that the precipitation
trend will be mainly negative, but the signal is not very robust across
GCMs, particularly toward the south. The country is located near the
dividing line between wetter and drier conditions for some parts of the
year. Hence the significance of the exact position of the divide between
wetter and drier conditions is speculative. Precipitation intensity will
show a downward trend, while the trend of dry day frequency continues
upward. Sea level rise will compare with global increase.
Images: Drier conditions but with some interannual variability.
Uruguay
Recent trends. Overall a modest temperature increase of 0.1°C per
decade has been observed during the past 25 years (1.5°C since 1901).
398 Christensen
Annual rainfall has significantly increased and sea level rise has occurred
at up to 3 mm/year.
Projections. See table 13.4 and figure 11.15 in WGII (IPCC 2007b) for
mean temperature and mean precipitation. Note that the precipitation
trend will be mainly positive throughout the year, but the signal is not
always robust across GCMs. The country is located near the dividing
line between wetter and drier conditions for some parts of the year.
Hence the significance of the exact position of the divide between wet-
ter and drier conditions is speculative. Precipitation intensity and dry
day frequency will both show an upward trend. Sea level rise will be
higher than global increases.
Images: Wetter conditions and increased risk for fast floods and
droughts.
Venezuela, R. B. De
Recent trends. Temperature increases at a rate of 0.1° to 0.2°C per
decade have been observed over the last 25 years (0.5° to 1.0°C since
1901, but information is only sparsely available). Annual rainfall shows
an increasing trend near the coast, while a negative trend is generally
observed in northern Amazonia. Sea level rise occurs at about 1.5 to
2 mm/year.
IBRD 37798
MAY 2010
- 0.2 - 0.1 0 0.1 0.2 meters
Source: Adapted from IPCC 2007b.
Note: Local sea level change (meters) due to ocean density and circulation change relative to the global average (positive values indicate greater local sea level change than global) during
the 21st century, calculated as the difference between averages for 2080–99 and those for 1980–99, as an ensemble mean over 16 AOGCMs forced with the Special Report on Emissions
399
Scenarios A1B scenario. Stippling denotes regions where the magnitude of the multimodel ensemble mean divided by the multimodel standard deviation exceeds 1.0.
400 Christensen
Projections. See table 13.4 and figure 11.15 in WGII (IPCC 2007b) for
mean temperature and mean precipitation. Note that the precipitation
trend will be mainly negative near the coast but positive toward the
Andes. Yet the signal is not very robust across GCMs. The country is
located near the dividing line between wetter and drier conditions for
some parts of the year. Hence the significance of the exact position of the
divide between wetter and drier conditions is speculative. Precipitation
intensity will show a downward trend in the north and increase toward
the Andes. The trend in dry day frequency continues upward. Sea level
rise will compare with global increases.
Images: Drier conditions but with large geographical variations.
References
IPCC (Intergovernmental Panel on Climate Change). 2007a. Climate Change
2007: The Physical Science Basis. Contribution of Working Group I to the
Fourth Assessment Report of the IPCC. Geneva: IPCC.
———. 2007b. Climate Change 2007: Impacts, Adaptation, and Vulnerability.
Contribution of Working Group II to the Fourth Assessment Report of the
IPCC. Geneva: IPCC.
———. 2007c. Synthesis Report: An Assessment of the Intergovernmental Panel on
Climate Change. Valencia, Spain, 12–17 November.
APPENDIX C
Livelihood assets
H
N P Transforming
Vulnerability Influence Livelihood Livelihood
structures
context & access strategies outcomes
F C and processes
S
Vulnerability Context
The “vulnerability context” refers to the external environment in which
people live. People’s livelihoods and the wider availability of assets are
largely affected by external trends, shocks, seasonality, and climatic vari-
ability. For the purpose of this book, the trends considered are gradual cli-
mate change and global warming, which may produce a wide range of
shocks (flooding, hurricanes, tornadoes, droughts), which in turn may
result in negative outcomes, such as economic and health conditions that
may lead to conflict, migration, and so on. These shocks will be indicative
of the current vulnerability of the populations examined. Climate change
may also affect the seasonality of climatic resources and increase climatic
variability, also affecting vulnerability.
Livelihood Assets
In this next step, the SLF considers the assets that individuals, households,
or communities have available. The assets considered include physical,
financial, human, social, cultural, and natural capital (table C.1).1
It is important to note that a single asset can generate multiple bene-
fits. For example, if a household has secure access to fertile land (natural
capital), it may also be well endowed with financial capital because it can
use the land not only directly for productive activities but also as collat-
eral (DFID 2004). Hence the SLF can aid in gauging the availability of
Methodology and Data 403
assets and determine how these interact with, or affect, one another, and
it can map which populations may be more affected by shocks than oth-
ers. It is understood that the more assets a person has available, the less
vulnerable he or she is.
DFID’s original SLF includes five types of livelihood assets, namely,
physical, natural, financial, social, and human capital. For this book, a
sixth type of livelihood asset, cultural capital, has been included. Field
research has revealed that the cultural dimension of livelihood strategies
and social institutions is particularly important for understanding and
describing the impacts of climate change and climatic variability experi-
enced by indigenous peoples.
“Cultural capital” can be defined as the knowledge, experience, and
connections that people have had throughout their lives, which enable
them to succeed better than someone without such a background.
Cultural capital is a sociological concept that has gained widespread
popularity since it was first articulated by Bourdieu.2 Cultural capital
acts as a social relation within a system of exchange, and the term is
extended to all the goods—material and symbolic—that present them-
selves as rare and worthy of being sought after in a particular social for-
mation. Thus, cultural capital acts as a social relation within a system
of exchange, which includes the accumulated cultural knowledge that
404 Andersen, Brisson, Pörtner, and Verner
Livelihood Strategies
The framework provides insight into how vulnerability and assets, as well
as structures and processes, influence livelihood strategies and how these
may be improved. The expansion of choice and value is important because
it provides people with opportunities for self-determination and the
Methodology and Data 405
Livelihood Outcomes
Finally, the framework examines how the livelihood strategies, given the
other factors, result in different livelihood outcomes and how the liveli-
hood outcomes feed back into available assets, creating either a virtuous
or vicious circle. Specifically, the framework helps to pinpoint entry
points for how to raise income, increase well-being, reduce vulnerability,
improve food security, and achieve more sustainable use of resources.
The study described in this book used the SLF to analyze
Quantitative Analyses
Weather-Related Disasters and Health Outcomes
Chapter 6, on the health impacts of climate change and variability,
includes analyses of the relationship between weather-related disasters and
health outcomes in Guatemala. The data used are described in table C.2.
406 Andersen, Brisson, Pörtner, and Verner
the error term and the shock variable. Which direction the bias will
take depends on how risk affects child health directly. If households
respond to higher risk by having more children, we might expect that
effect to be negative for the anthropometric measures because of
closer spacing of children and a more binding resource constraint.
Obviously, there is a positive correlation between risk and the number
of shocks that occur in a department, which means that the bias, in this
case, will be downward.
The household- and individual-level explanatory variables used are the
sex of the child; the age of the child in months at the time of the survey;
the age of the mother at the time of the child’s birth; her education and
literacy levels; the father’s education level; and the ethnicity of the child.
For both parental age and education levels, the squares of the variables are
also included. Unfortunately, the data do not include direct information
on land ownership, and so a dummy variable is included as a proxy: this
takes the value of 1 if either the mother or the father responds that they
work on their own or family land, and zero otherwise.
expectancy data are not available) for five countries in Latin America and
the Caribbean (Brazil, Mexico, Chile, Peru, and Bolivia) to estimate the
likely relationships between climate, on one hand, and life expectancy/child
mortality (chapter 6) and poverty and inequality (chapter 7) on the other.
Both average annual temperature and its square are included in the
regression model, and so is rainfall squared. The regressions control for
other factors that may affect the level of development but are likely to be
insensitive to climate change in the short run. The key control variables
are education level and urbanization level.
Apart from income level as a measure of development, life
expectancy is also used. The life expectancy regression takes the same
form as the income regressions, except that the natural logarithm is
not applied to the dependent variable. All regressions are weighted
ordinary least squares regressions, where the weights consist of the
population size in each municipality.
Once estimated, the country-specific relationships between climate
and income and climate and life expectancy are used to gauge the
likely effects of climate change for each municipality in the countries
under investigation. The effects of the climate change that has taken
place during the last 50 years, as documented by local temperature
records, are analyzed, as are the effects of expected future climate
change over the next 50 years, as projected by IPCC’s Atmosphere-
Ocean General Circulation Models. Evaluating the effects for each
municipality enables us to assess whether climate change is contribut-
ing to increased inequality and poverty, by testing whether initially
poorer municipalities are more adversely affected by climate change
than are initially richer municipalities.
+ βˆ 4 ⋅ raini2,NCC + ∑ αˆ X
j =1
j j ,i + εˆi
where the index i refers to municipality i; temp and rain are the temper-
ature and rainfall variables; the bˆ s are the estimated coefficients on the
temperature and rainfall variables; the Xj s are the remaining j explanatory
variables including the constant term; the α̂ j s are the coefficient to these
variables; and ε̂ i are the estimated error terms for each municipality.
Equivalently, the level of health under the assumption of climate
change can be written as follows:
ˆ ˆ ˆ
i ,CC = β1 ⋅ tempi ,CC + β 2 ⋅ tempi ,CC
2
health
k
where the only differences are the temperature and rainfall variables. The
control variables are held constant, so as to isolate the effects of climate
change and variability.
The difference in life expectancy that can be directly attributed to cli-
mate change and variability can be found as the difference between the
two scenarios:
ˆ
Δ CC health ˆ ˆ
i = healthi ,CC − healthi , NCC
ln(Yi ,NCC ) = αˆ + βˆ1⋅ tempi ,NCC + βˆ2 ⋅ tempi2,NCC + βˆ 3 ⋅ raini ,NCC + βˆ 4 ⋅ raini2,NCC
+ βˆ5⋅ edui + βˆ6 ⋅ urbi + βˆ7 ⋅ urbi2 + εˆ i
,
where the only differences are the temperature and rainfall levels. All
other variables are held constant in the simulation. The ratio of Climate
Change Income to No Climate Change Income can then be written as
follows:
Yˆi ,CC
Δ CCYˆi =
Yˆi , NCC
=
{
exp βˆ1 ⋅ tempi ,CC + βˆ 2 ⋅ tempi2,CC + βˆ 3 ⋅ raini ,CC + βˆ 4 ⋅ raini2,CC }
{
exp βˆ1 ⋅ tempi ,NCC + βˆ 2 ⋅ tempi2,NCC + βˆ 3 ⋅ raini ,NCC + βˆ 4 ⋅ raini2,NCC }
After estimating this ratio for each municipality, it is easy to calculate
the percentage change in income that can be attributed to climate
change.
412 Andersen, Brisson, Pörtner, and Verner
⎛ ⎞
⎜ ⎟
⎛ Mii ⎞ ⎜ Mij ⎟
mij = ⎜1 − ⋅
⎝ pobi ⎟⎠ ⎜⎜ n ⎟
⎜ ∑
⎝ k =1,k ≠ i
Mik ⎟
⎟
⎠
where the first part is the proportion of persons who decide to migrate
from municipality i and the second part is the probability that they
choose destination j, given that they have decided to migrate. This speci-
fication reflects the two step process of migration: first people decide
whether to migrate or not, and then they decide where to go.
Given that, three different models can be estimated to analyze the
determinants of migration:
Model 1 (out-migration):
Methodology and Data 413
Mij
1− = α + β ′( X i )
pobi
Mij
n
= α + β ′( dij , X j ).
∑
k =1,k ≠ i
Mik
Notes
1. The shape of the asset hexagon (see figure C.1) can be used to show graphi-
cally the variation in people’s access to assets. The center point of the hexagon,
where the lines meet, represents zero access to assets. The greater the distance
from the center, the greater is the access to any given asset. Thus, on this basis,
in principle, differently shaped hexagons can be drawn for different commu-
nities or social groups within communities.
2. Pierre Bourdieu and Jean-Claude Passeron first used the term in Cultural
Reproduction and Social Reproduction (Bourdieu 1973). In that work, Bourdieu
attempted to explain differences in educational outcomes in France during
the 1960s. The concept has since been elaborated and developed in terms of
414 Andersen, Brisson, Pörtner, and Verner
other types of capital in The Forms of Capital (Bourdieu 1983, 1986) and in
terms of higher education in, for instance, in The State Nobility (Bourdieu, de
Saint Martin, and Clough 1996).
3. There are 22 departments in Guatemala, with a total of 331 municipalities.
4. Horowitz looked at a 2°F increase in temperature, which is equivalent to a
1.1°C increase.
References
Bourdieu, Pierre. 1973. “Cultural Reproduction and Social Reproduction.” In
Knowledge, Education, and Cultural Change, ed. Richard Brown. London:
Willmer Brothers Ltd.
———. 1983, 1986. “The Forms of Capital.” Trans. by Richard Nice. In Handbook
of Theory and Research for the Sociology of Education, ed. J. G. Richardson.
241–58. New York: Greenwood Press.
Bourdieu, Pierre, Monique de Saint Martin, and Laurette C. Clough. 1996. The
State Nobility. Stanford, CA: Stanford University Press.
DFID (Department for International Development, U.K.). 2004. Climate Change in
Latin America. Key Fact Sheet No. 12. Key Fact Sheets on Climate Change and
Poverty. http://www.dfid.gov.uk/pubs/files/climatechange/keysheetsindex.asp.
Horowitz, J. K. 2006. “The Income-Temperature Relationship in a Cross-Section
of Countries and Its Implications for Global Warming.” Department of
Agricultural and Resource Economics, University of Maryland, Submitted
manuscript, July. http://faculty.arec.umd.edu/jhorowitz/Income-Temp-i.pdf.
Pörtner, Claus C. 2008. “Natural Hazards and Child Health.” Department of
Economics, University of Washington. http://faculty.washington.edu/
cportner/nathaz.pdf.
UNICEF. 2000. “Desastres Naturales Y Zonas De Riesgo En Guatemala.”
Unpublished paper, UNICEF.
WHO (World Health Organization). 2006. “WHO Child Growth Standards.”
World Health Organization, Geneva. http://www.who.int/childgrowth/
standards/en/.
Index
Boxes, figures, notes, and tables are indicated by b, f, n, or t following the page number.
415
416 Index
natural disaster impact on, 69, 81–82b, surface runoff and waterborne
85–86, 87, 88n12 diseases, 179
policy implications, 5–6, 6f Suriname
research recommendations, 334 malaria in, 167
rural poverty and, 97 water scarcity, 38
in SLF, 13, 14t surveillance systems (health), 188, 191n14
urban poor and, 147, 148 sustainable livelihoods framework (SLF)
vulnerability reduction and, 331–32 agrarian livelihoods, 93
youth and, 244n10 coastal communities, 130
social memory, 276–77, 278, 287 health and, 169
socioeconomic context livelihood assets, 13–14, 14t
of conflict, 234–38, 236t, 244n9 livelihood strategies and outcomes, 15
of dengue fever outbreaks, 178 migration and, 197, 204
of waterborne diseases, 180 natural disasters and, 64
soil degradation, 98–100, 99f, overview, 12–15, 13b
225, 227, 315 social capital in, 269
Sokona, Y., 69 transformational structures
South America. See also specific countries and processes, 14
dengue fever in, 176, 322 vulnerability context, 13
ENSO impact on, 67, 67f water scarcity and, 53
floods in, 65
natural disaster patterns in, 70–71, 71t
T
Southern Cone region
economic integration in, 222 tariffs, water, 47, 56n22
malaria in, 175 technical capacity, 330–31
Soysa, I., 225 technical committees for groundwater
Special Report on Emissions Scenarios (COTAS, Mexico), 55–56n19
(SRES, IPCC), 39 temperature change
Sri Lanka, tourism losses related to air quality and, 171
tsunami in, 143 disease and, 170, 172, 179
staple crops, 17n5, 206 health impacts of, 322–23
state capacity and conflict, 224, income levels and, 249, 250f, 252,
237, 326 253, 254f, 257, 261
Stern Review on the Economics of Climate migration and, 208, 211, 211f,
Change (Stern), 6–7, 16n4 216f, 218n6
storage projection assumptions, 9
crop storage, 110 social impacts of, 313–15
water tanks, 40b, 179 tourism impact of, 140–41
stormwater drainage systems, 35 vector-borne diseases and, 172
Strauss, J., 181, 190n10 waterborne diseases and, 179
structural adjustment policies, temporary hunger, 80, 314
100–101, 115–17 temporary migration, 200
subsidiarity principle, 42 TFESSD (Trust Fund for Environmentally
subsistence farming and Socially Sustainable
adaptive capacity and, 77 Development), 299n19
coping capacity of, 273 Thailand, tourism losses related
droughts and, 284 to tsunami in, 143
natural disaster impact on, 81–82b, 83 thermal power generation, 49
rural poverty and, 96, 96t Thomas, D., 181, 190n10
social capital and, 88n12 Tikjøb, Sanne, 305
subtropical high-pressure zone, 65 timber industry, 100
sun and surf tourism, 158–59n15 Tol, R. S.J., 140
430 Index
In this important and provocative volume, Dorte Verner and her colleagues provide an
expansive treatment of climate change and its many effects, especially on the global poor.
—David Lee, Professor, Cornell University
This book is bound to become the defining analysis of climate change’s implications for
poverty and social cohesion. . . . A key guide for policy makers.
—Daniel Cohen, Professor and Vice President, Paris School of Economics
This much-welcome overview provides guidelines and suggests priorities for designing
and implementing suitable adaptation measures.
—Anthony Hall, Professor, London School of Economics
It is no longer possible to prevent damaging climate change. If used properly, this book will
save many lives.
—Robert Waldmann, Professor, University of Rome
This book is simply a must for all those concerned by climate change.
—Javier Santiso, Professor of Economics, ESADE Business School
This is first time to have a comprehensive social assessment of climate change and it will
become a standard reference.
—Shengen Fan, Director General, International Food Policy Research Institute
This book presents evidence that we must improve our efforts on resilience and adaptation
measures to counter the consequences of climate change on the most vulnerable population
groups.
—Søren Pind, Minister for Development Cooperation, Denmark
SKU 18238